gpuark-benchmarks.csv

gpu_id,test_name,score,unit,source__name
10819,Hashcat: Benchmark: SHA-256,9853600000.0,H/s,OpenBenchmarking: Hashcat
3379,Hashcat: Benchmark: NTLM,10387833333.0,H/s,OpenBenchmarking: Hashcat
3379,Hashcat: Benchmark: SHA-512,250700000.0,H/s,OpenBenchmarking: Hashcat
3379,Hashcat: Benchmark: 7-Zip,80162.0,H/s,OpenBenchmarking: Hashcat
3379,Hashcat: Benchmark: LastPass,3006.0,H/s,OpenBenchmarking: Hashcat
10335,Llama.cpp: Backend: CPU BLAS - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Prompt Processing 2048,87.46,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,15.67,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 512,357.2,Tokens Per Second,OpenBenchmarking: Llama.cpp
3379,Hashcat: Benchmark: SHA1,1965200000.0,H/s,OpenBenchmarking: Hashcat
10335,vLLM CPU: Test: deepseek-moe-16b-chat Latency,23.09,99p Latency - Seconds,OpenBenchmarking: Llama.cpp
3379,Hashcat: Benchmark: MD5,6116766667.0,H/s,OpenBenchmarking: Hashcat
10335,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024,38.24,Tokens Per Second,OpenBenchmarking: Llama.cpp
3379,Hashcat: Benchmark: SHA-256,450900000.0,H/s,OpenBenchmarking: Hashcat
3379,Hashcat: Benchmark: TrueCrypt RIPEMD160 + XTS,69255.0,H/s,OpenBenchmarking: Hashcat
3379,Hashcat: Benchmark: Kerberos 5 + etype 23 + TGS-REP,93845167.0,H/s,OpenBenchmarking: Hashcat
3379,Hashcat: Benchmark: LUKS v2 argon2id + SHA-256 + AES,2.0,H/s,OpenBenchmarking: Hashcat
1094,Hashcat: Benchmark: WPA-PBKDF2-PMKID+EAPOL,240200.0,H/s,OpenBenchmarking: Hashcat
10335,Hashcat: Benchmark: SHA-256,604199142.0,H/s,OpenBenchmarking: Hashcat
10335,vLLM CPU: Test: Hermes-3-Llama-3.2-3B Latency,9.92,99p Latency - Seconds,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512,225.25,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Prompt Processing 1024,84.71,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512,1726.09,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512,90.28,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,26.17,Tokens Per Second,OpenBenchmarking: Llama.cpp
3379,Hashcat: Benchmark: bcrypt,5430.0,H/s,OpenBenchmarking: Hashcat
10910,vkpeak: fp32-scalar,34789.41,GFLOPS,OpenBenchmarking: Hashcat
10910,vkpeak: fp32-vec4,34412.58,GFLOPS,OpenBenchmarking: Hashcat
10910,vkpeak: fp16-scalar,25965.34,GFLOPS,OpenBenchmarking: Hashcat
10910,vkpeak: fp16-vec4,25822.58,GFLOPS,OpenBenchmarking: Hashcat
10910,vkpeak: fp16-matrix,205769.65,GFLOPS,OpenBenchmarking: Hashcat
10761,Hashcat: Benchmark: NTLM,179533333333.0,H/s,OpenBenchmarking: Hashcat
10761,Hashcat: Benchmark: SHA1,38457866667.0,H/s,OpenBenchmarking: Hashcat
10761,Hashcat: Benchmark: 7-Zip,1767267.0,H/s,OpenBenchmarking: Hashcat
10761,Hashcat: Benchmark: bcrypt,133400.0,H/s,OpenBenchmarking: Hashcat
10761,Hashcat: Benchmark: SHA-256,13393933333.0,H/s,OpenBenchmarking: Hashcat
10761,Hashcat: Benchmark: SHA-512,4443333333.0,H/s,OpenBenchmarking: Hashcat
10761,Hashcat: Benchmark: LastPass,55862.0,H/s,OpenBenchmarking: Hashcat
10761,Hashcat: Benchmark: WPA-PBKDF2-PMKID+EAPOL,1761767.0,H/s,OpenBenchmarking: Hashcat
10761,Hashcat: Benchmark: TrueCrypt RIPEMD160 + XTS,1281400.0,H/s,OpenBenchmarking: Hashcat
10761,Hashcat: Benchmark: Kerberos 5 + etype 23 + TGS-REP,2110933333.0,H/s,OpenBenchmarking: Hashcat
10761,Hashcat: Benchmark: LUKS v2 argon2id + SHA-256 + AES,12.0,H/s,OpenBenchmarking: Hashcat
5057,LLM Inference: LLM Perf: Phi-3-mini-4k-instruct (awq),24.443189211775355,tokens/s,LLM Inference
10938,LLM Inference: Geekbench AI: Geekbench AI Quantized (int8),33825.0,points,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-7B (gptq),28.810004080992993,tokens/s,LLM Inference
1814,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),9.49,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Meta-Llama-3-8B-Instruct (awq),22.27493301865801,tokens/s,LLM Inference
10910,Hashcat: Benchmark: SHA-512,4112433333.0,H/s,OpenBenchmarking: Hashcat
10910,Hashcat: Benchmark: LastPass,47630.0,H/s,OpenBenchmarking: Hashcat
10910,Hashcat: Benchmark: WPA-PBKDF2-PMKID+EAPOL,1400167.0,H/s,OpenBenchmarking: Hashcat
10910,Hashcat: Benchmark: TrueCrypt RIPEMD160 + XTS,1159267.0,H/s,OpenBenchmarking: Hashcat
10910,Hashcat: Benchmark: LUKS v2 argon2id + SHA-256 + AES,13.0,H/s,OpenBenchmarking: Hashcat
10910,cl-mem: Benchmark: Read,793.5,GB/s,OpenBenchmarking: Hashcat
10910,cl-mem: Benchmark: Write,775.8,GB/s,OpenBenchmarking: Hashcat
10910,"NAMD CUDA: Input: ATPase with 327,506 Atoms",11.88743,ns/day,OpenBenchmarking: Hashcat
10910,"NAMD CUDA: Input: STMV with 1,066,628 Atoms",3.51814,ns/day,OpenBenchmarking: Hashcat
10910,VkResample: Upscale: 2x - Precision: Single,11.643,ms,OpenBenchmarking: Hashcat
10910,FAHBench,389.0836,Ns Per Day,OpenBenchmarking: Hashcat
10910,clpeak: OpenCL Test: Double-Precision Compute,812.66,GFLOPS,OpenBenchmarking: Hashcat
10910,ViennaCL: Test: OpenCL BLAS - sAXPY,449.0,GB/s,OpenBenchmarking: Hashcat
10910,ViennaCL: Test: OpenCL BLAS - sDOT,367.0,GB/s,OpenBenchmarking: Hashcat
10910,ViennaCL: Test: OpenCL BLAS - dCOPY,546.0,GB/s,OpenBenchmarking: Hashcat
10910,ViennaCL: Test: OpenCL BLAS - dAXPY,683.0,GB/s,OpenBenchmarking: Hashcat
10910,ViennaCL: Test: OpenCL BLAS - dDOT,626.0,GB/s,OpenBenchmarking: Hashcat
10910,ViennaCL: Test: OpenCL BLAS - dGEMV-N,184.0,GB/s,OpenBenchmarking: Hashcat
10910,ViennaCL: Test: OpenCL BLAS - dGEMV-T,345.0,GB/s,OpenBenchmarking: Hashcat
10910,ViennaCL: Test: OpenCL BLAS - dGEMM-NN,664.0,GFLOPs/s,OpenBenchmarking: Hashcat
10910,ViennaCL: Test: OpenCL BLAS - dGEMM-NT,711.0,GFLOPs/s,OpenBenchmarking: Hashcat
10910,ViennaCL: Test: OpenCL BLAS - dGEMM-TN,712.0,GFLOPs/s,OpenBenchmarking: Hashcat
10910,ViennaCL: Test: OpenCL BLAS - dGEMM-TT,776.0,GFLOPs/s,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: mobilenet,25.89,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2,5.94,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3,4.05,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: shufflenet-v2,3.65,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: mnasnet,4.97,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: efficientnet-b0,7.94,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: blazeface,2.32,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: googlenet,8.95,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: vgg16,37.44,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: resnet18,6.25,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: alexnet,6.21,ms,OpenBenchmarking: Hashcat
11889,vkpeak: fp32-vec4,34412.58,GFLOPS,OpenBenchmarking: Hashcat
10938,LLM Inference: Geekbench AI: Geekbench AI Single Precision (fp32),45417.0,points,LLM Inference
11889,vkpeak: fp16-scalar,25965.34,GFLOPS,OpenBenchmarking: Hashcat
11889,vkpeak: fp16-vec4,25822.58,GFLOPS,OpenBenchmarking: Hashcat
11889,vkpeak: fp16-matrix,205769.65,GFLOPS,OpenBenchmarking: Hashcat
10938,LLM Inference: Geekbench AI: Geekbench AI Half Precision (fp16),66366.0,points,LLM Inference
10335,OpenRadioss: Model: Rubber O-Ring Seal Installation,50.75,Seconds,OpenBenchmarking: Llama.cpp
10335,QMCPACK: Input: H4_ae,10.72,Total Execution Time - Seconds,OpenBenchmarking: Llama.cpp
10335,OpenRadioss: Model: Bird Strike on Windshield,105.37,Seconds,OpenBenchmarking: Llama.cpp
10335,"ClickHouse: 100M Rows Hits Dataset, Latency, Second Run",861.82,"Queries Per Minute, Geo Mean",OpenBenchmarking: Llama.cpp
10335,OpenRadioss: Model: Bumper Beam,77.15,Seconds,OpenBenchmarking: Llama.cpp
11889,vkpeak: fp64-vec4,812.64,GFLOPS,OpenBenchmarking: Hashcat
10335,QMCPACK: Input: Li2_STO_ae,97.311,Total Execution Time - Seconds,OpenBenchmarking: Llama.cpp
10335,"ClickHouse: 100M Rows Hits Dataset, Throughput",245.09,"Queries Per Minute, Geo Mean",OpenBenchmarking: Llama.cpp
11889,vkpeak: int16-vec4,24589.36,GIOPS,OpenBenchmarking: Hashcat
11889,vkpeak: int8-dotprod,32808.08,GIOPS,OpenBenchmarking: Hashcat
11889,vkpeak: int8-matrix,409762.38,GIOPS,OpenBenchmarking: Hashcat
11889,vkpeak: int32-scalar,25964.43,GIOPS,OpenBenchmarking: Hashcat
11889,vkpeak: bf16-matrix,101289.72,GFLOPS,OpenBenchmarking: Hashcat
11889,RealSR-NCNN: Scale: 4x - TAA: No,5.589,Seconds,OpenBenchmarking: Hashcat
11889,vkpeak: fp64-scalar,813.8,GFLOPS,OpenBenchmarking: Hashcat
11889,Waifu2x-NCNN Vulkan: Scale: 2x - Denoise: 3 - TAA: No,5.878,Seconds,OpenBenchmarking: Hashcat
11889,Hashcat: Benchmark: MD5,90224866667.0,H/s,OpenBenchmarking: Hashcat
11889,Hashcat: Benchmark: NTLM,139566666667.0,H/s,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: resnet50,14.33,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: yolov4-tiny,29.92,ms,OpenBenchmarking: Hashcat
10910,Blender: Blend File: Fishy Cat - Compute: NVIDIA OptiX,8.45,Seconds,OpenBenchmarking: Hashcat
10910,Blender: Blend File: Barbershop - Compute: NVIDIA OptiX,46.33,Seconds,OpenBenchmarking: Hashcat
10910,MandelGPU: OpenCL Device: GPU,831796323.4,Samples/sec,OpenBenchmarking: Hashcat
10910,NeatBench: Acceleration: GPU,29.97,FPS,OpenBenchmarking: Hashcat
10761,Hashcat: Benchmark: MD5,105100000000.0,H/s,OpenBenchmarking: Hashcat
10335,vLLM CPU: Test: Hermes-3-Llama-3.2-3B Throughput,198.31,tokens/s,OpenBenchmarking: Llama.cpp
11889,vkpeak: int32-vec4,24593.47,GIOPS,OpenBenchmarking: Hashcat
11889,Hashcat: Benchmark: SHA1,27785466667.0,H/s,OpenBenchmarking: Hashcat
11889,Hashcat: Benchmark: 7-Zip,1441767.0,H/s,OpenBenchmarking: Hashcat
11889,Hashcat: Benchmark: bcrypt,128000.0,H/s,OpenBenchmarking: Hashcat
10335,"ClickHouse: 100M Rows Hits Dataset, Latency, Third Run",880.11,"Queries Per Minute, Geo Mean",OpenBenchmarking: Llama.cpp
10335,Redis: Test: GET - Parallel Connections: 50,4161464.75,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,CloverLeaf: Input: clover_bm64_short,95.4,Seconds,OpenBenchmarking: Llama.cpp
3667,Hashcat: Benchmark: WPA-PBKDF2-PMKID+EAPOL,651833.0,H/s,OpenBenchmarking: Hashcat
3667,Hashcat: Benchmark: LUKS v2 argon2id + SHA-256 + AES,6.0,H/s,OpenBenchmarking: Hashcat
11889,vkpeak: int16-scalar,25955.42,GIOPS,OpenBenchmarking: Hashcat
10335,OpenRadioss: Model: Cell Phone Drop Test,21.69,Seconds,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: HSET - Parallel Connections: 800,669613.0,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,81.35,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: GET - Parallel Connections: 800,1127472.0,Requests Per Second,OpenBenchmarking: Llama.cpp
11889,Hashcat: Benchmark: SHA-256,11578933333.0,H/s,OpenBenchmarking: Hashcat
10913,LLM Inference: MLPerf: deepseek-r1 (fp8),1113.10375,tokens/s,LLM Inference
10335,Valkey: Test: INCR - Parallel Connections: 500,1239310.88,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: SET - Parallel Connections: 500,652145.56,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: SPOP - Parallel Connections: 800,1180609.62,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: SET - Parallel Connections: 800,662883.81,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: INCR - Parallel Connections: 800,1312439.25,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512,170.79,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,15.31,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: LPOP - Parallel Connections: 500,1395556.62,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512,64.69,Tokens Per Second,OpenBenchmarking: Llama.cpp
11889,Hashcat: Benchmark: SHA-512,4112433333.0,H/s,OpenBenchmarking: Hashcat
10335,Valkey: Test: HSET - Parallel Connections: 500,662954.12,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048,79.41,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512,78.81,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: LPOP - Parallel Connections: 800,1405047.0,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128,39.96,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: SADD - Parallel Connections: 500,997804.81,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: SADD - Parallel Connections: 800,1007698.81,Requests Per Second,OpenBenchmarking: Llama.cpp
11889,Hashcat: Benchmark: LastPass,47630.0,H/s,OpenBenchmarking: Hashcat
11889,Hashcat: Benchmark: WPA-PBKDF2-PMKID+EAPOL,1400167.0,H/s,OpenBenchmarking: Hashcat
11889,Hashcat: Benchmark: TrueCrypt RIPEMD160 + XTS,1159267.0,H/s,OpenBenchmarking: Hashcat
11889,Hashcat: Benchmark: Kerberos 5 + etype 23 + TGS-REP,1804800000.0,H/s,OpenBenchmarking: Hashcat
11889,Hashcat: Benchmark: LUKS v2 argon2id + SHA-256 + AES,13.0,H/s,OpenBenchmarking: Hashcat
11889,cl-mem: Benchmark: Read,793.5,GB/s,OpenBenchmarking: Hashcat
11889,cl-mem: Benchmark: Write,775.8,GB/s,OpenBenchmarking: Hashcat
11889,"NAMD CUDA: Input: ATPase with 327,506 Atoms",11.88743,ns/day,OpenBenchmarking: Hashcat
11889,"NAMD CUDA: Input: STMV with 1,066,628 Atoms",3.51814,ns/day,OpenBenchmarking: Hashcat
11889,VkResample: Upscale: 2x - Precision: Single,11.643,ms,OpenBenchmarking: Hashcat
11889,FAHBench,389.0836,Ns Per Day,OpenBenchmarking: Hashcat
11889,clpeak: OpenCL Test: Double-Precision Compute,812.66,GFLOPS,OpenBenchmarking: Hashcat
11889,Rodinia: Test: OpenMP HotSpot3D,39.376,Seconds,OpenBenchmarking: Hashcat
11889,Rodinia: Test: OpenMP Leukocyte,31.639,Seconds,OpenBenchmarking: Hashcat
11889,Rodinia: Test: OpenMP CFD Solver,9.038,Seconds,OpenBenchmarking: Hashcat
11889,Rodinia: Test: OpenMP Streamcluster,11.165,Seconds,OpenBenchmarking: Hashcat
10335,Valkey: Test: INCR - Parallel Connections: 50,1416631.38,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: HSET - Parallel Connections: 50,729128.75,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048,84.58,Tokens Per Second,OpenBenchmarking: Llama.cpp
11889,FinanceBench: Benchmark: Repo OpenMP,16807.688802,ms,OpenBenchmarking: Hashcat
10335,Llama.cpp: Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 1024,321.01,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128,16.47,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048,38.24,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 2048,303.51,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048,1030.45,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512,1114.99,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512,86.86,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024,86.69,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,25.79,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512,117.26,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 1024,84.72,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,14.04,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128,78.56,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,1098.57,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Prompt Processing 512,88.29,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,116.44,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048,113.97,Tokens Per Second,OpenBenchmarking: Llama.cpp
11889,FinanceBench: Benchmark: Bonds OpenMP,28291.983724,ms,OpenBenchmarking: Hashcat
11889,FinanceBench: Benchmark: Monte-Carlo OpenCL,102.072001,ms,OpenBenchmarking: Hashcat
11889,FinanceBench: Benchmark: Black-Scholes OpenCL,4.174,ms,OpenBenchmarking: Hashcat
11889,ViennaCL: Test: OpenCL BLAS - sCOPY,334.0,GB/s,OpenBenchmarking: Hashcat
11889,ViennaCL: Test: OpenCL BLAS - sAXPY,449.0,GB/s,OpenBenchmarking: Hashcat
11889,ViennaCL: Test: OpenCL BLAS - sDOT,367.0,GB/s,OpenBenchmarking: Hashcat
11889,ViennaCL: Test: OpenCL BLAS - dCOPY,546.0,GB/s,OpenBenchmarking: Hashcat
11889,ViennaCL: Test: OpenCL BLAS - dAXPY,683.0,GB/s,OpenBenchmarking: Hashcat
11889,ViennaCL: Test: OpenCL BLAS - dDOT,626.0,GB/s,OpenBenchmarking: Hashcat
11889,ViennaCL: Test: OpenCL BLAS - dGEMV-N,184.0,GB/s,OpenBenchmarking: Hashcat
11889,ViennaCL: Test: OpenCL BLAS - dGEMV-T,345.0,GB/s,OpenBenchmarking: Hashcat
11889,ViennaCL: Test: OpenCL BLAS - dGEMM-NN,664.0,GFLOPs/s,OpenBenchmarking: Hashcat
11889,ViennaCL: Test: OpenCL BLAS - dGEMM-NT,711.0,GFLOPs/s,OpenBenchmarking: Hashcat
11889,ViennaCL: Test: OpenCL BLAS - dGEMM-TN,712.0,GFLOPs/s,OpenBenchmarking: Hashcat
11889,ViennaCL: Test: OpenCL BLAS - dGEMM-TT,776.0,GFLOPs/s,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU - Model: mobilenet,25.89,ms,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU-v2-v2 - Model: mobilenet-v2,5.94,ms,OpenBenchmarking: Hashcat
11889,Rodinia: Test: OpenCL Particle Filter,2.921,Seconds,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU - Model: shufflenet-v2,3.65,ms,OpenBenchmarking: Hashcat
9882,Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128,46.2,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512,953.45,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024,30.14,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048,124.31,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048,1278.18,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512,30.19,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,1337.74,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128,19.7,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128,37.59,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048,212.58,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128,71.45,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024,914.59,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048,28.95,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024,1721.44,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128,59.44,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048,83.29,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,14.25,Tokens Per Second,OpenBenchmarking: Llama.cpp
11889,NCNN: Target: Vulkan GPU - Model: mnasnet,4.97,ms,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU - Model: efficientnet-b0,7.94,ms,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU - Model: blazeface,2.32,ms,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU - Model: googlenet,8.95,ms,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU - Model: vgg16,37.44,ms,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU - Model: resnet18,6.25,ms,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU - Model: alexnet,6.21,ms,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU - Model: resnet50,14.33,ms,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU - Model: yolov4-tiny,29.92,ms,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU - Model: squeezenet_ssd,9.84,ms,OpenBenchmarking: Hashcat
9882,Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048,834.67,Tokens Per Second,OpenBenchmarking: Llama.cpp
11889,NCNN: Target: Vulkan GPU - Model: regnety_400m,23.62,ms,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU - Model: vision_transformer,77.23,ms,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU - Model: FastestDet,5.21,ms,OpenBenchmarking: Hashcat
11889,Blender: Blend File: Fishy Cat - Compute: NVIDIA CUDA,20.6,Seconds,OpenBenchmarking: Hashcat
11889,Blender: Blend File: Junkshop - Compute: NVIDIA OptiX,8.28,Seconds,OpenBenchmarking: Hashcat
11889,Blender: Blend File: Barbershop - Compute: NVIDIA CUDA,78.09,Seconds,OpenBenchmarking: Hashcat
11889,NCNN: Target: Vulkan GPU-v3-v3 - Model: mobilenet-v3,4.05,ms,OpenBenchmarking: Hashcat
10810,LLM Inference: MLPerf: llama2-70b-99.9 (fp8),1945.775,tokens/s,LLM Inference
10913,LLM Inference: MLPerf: llama2-70b-99.9 (fp8),2444.3375,tokens/s,LLM Inference
10810,LLM Inference: MLPerf: mixtral-8x7b (fp8),6975.0375,tokens/s,LLM Inference
10335,Llama.cpp: Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 512,84.74,Tokens Per Second,OpenBenchmarking: Llama.cpp
3860,LLM Inference: MLPerf: llama2-70b-99.9 (UNSET),1982.9625,tokens/s,LLM Inference
10810,LLM Inference: MLPerf: llama2-70b-99.9 (Dummy),3596.275,tokens/s,LLM Inference
10913,LLM Inference: MLPerf: llama2-70b-99.9 (fp16),3552.625,tokens/s,LLM Inference
3860,LLM Inference: MLPerf: llama3.1-405b (fp16),34.831125,tokens/s,LLM Inference
10913,LLM Inference: MLPerf: mixtral-8x7b (fp8),7131.8,tokens/s,LLM Inference
3860,LLM Inference: MLPerf: mixtral-8x7b (fp8),5975.2375,tokens/s,LLM Inference
11342,LLM Inference: MLPerf: llama3.1-405b (mxfp4),252.38,tokens/s,LLM Inference
9882,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048,113.19,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,26.29,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024,1079.59,Tokens Per Second,OpenBenchmarking: Llama.cpp
11342,LLM Inference: MLPerf: llama2-70b-99 (mxfp4),10128.875,tokens/s,LLM Inference
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512,237.79,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024,236.95,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048,964.38,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048,230.31,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128,27.44,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024,120.17,Tokens Per Second,OpenBenchmarking: Llama.cpp
3860,LLM Inference: MLPerf: llama2-70b-99.9 (fp8),1056.8925,tokens/s,LLM Inference
10913,LLM Inference: MLPerf: llama3.1-8b (fp8),5161.375,tokens/s,LLM Inference
3860,LLM Inference: MLPerf: llama3.1-405b (UNSET),30.431125,tokens/s,LLM Inference
10913,LLM Inference: MLPerf: gptj-99.9 (fp8),2562.725,tokens/s,LLM Inference
11342,LLM Inference: MLPerf: llama2-70b-99.9 (mxfp4),10128.875,tokens/s,LLM Inference
3860,LLM Inference: MLPerf: llama2-70b-99.9 (fp16),1739.525,tokens/s,LLM Inference
10810,LLM Inference: MLPerf: llama2-70b-99 (Dummy),3596.275,tokens/s,LLM Inference
10913,LLM Inference: MLPerf: llama2-70b-99.9 (UNSET),3534.05,tokens/s,LLM Inference
10913,LLM Inference: MLPerf: llama2-70b-99 (fp8),2444.3375,tokens/s,LLM Inference
1819,LLM Inference: llama.cpp Bench: llama 7B Q4_0 (Q4_0),14.27,tokens/s,LLM Inference
11889,Blender: Blend File: Fishy Cat - Compute: NVIDIA OptiX,8.45,Seconds,OpenBenchmarking: Hashcat
11889,Blender: Blend File: Barbershop - Compute: NVIDIA OptiX,46.33,Seconds,OpenBenchmarking: Hashcat
11889,MandelGPU: OpenCL Device: GPU,831796323.4,Samples/sec,OpenBenchmarking: Hashcat
11889,NeatBench: Acceleration: GPU,29.97,FPS,OpenBenchmarking: Hashcat
11743,Hashcat: Benchmark: MD5,105100000000.0,H/s,OpenBenchmarking: Hashcat
11743,Hashcat: Benchmark: NTLM,179533333333.0,H/s,OpenBenchmarking: Hashcat
11743,Hashcat: Benchmark: SHA1,38457866667.0,H/s,OpenBenchmarking: Hashcat
11743,Hashcat: Benchmark: 7-Zip,1767267.0,H/s,OpenBenchmarking: Hashcat
11743,Hashcat: Benchmark: bcrypt,133400.0,H/s,OpenBenchmarking: Hashcat
11743,Hashcat: Benchmark: SHA-256,13393933333.0,H/s,OpenBenchmarking: Hashcat
11743,Hashcat: Benchmark: SHA-512,4443333333.0,H/s,OpenBenchmarking: Hashcat
11743,Hashcat: Benchmark: LastPass,55862.0,H/s,OpenBenchmarking: Hashcat
10810,LLM Inference: MLPerf: llama2-70b-99 (fp8),1945.775,tokens/s,LLM Inference
11445,LLM Inference: llama.cpp Bench: llama 7B Q4_0 (Q4_0),287.29,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-70m (unquantized),159.81332717353104,tokens/s,LLM Inference
11557,LLM Inference: MLPerf: llama2-70b-99.9 (fp8),2106.075,tokens/s,LLM Inference
199,LLM Inference: MLPerf: llama2-70b-99.9 (fp8),1222.48,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: distilgpt2 (gptq),187.98064957142236,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt2 (unquantized),101.1161948589665,tokens/s,LLM Inference
196,LLM Inference: MLPerf: mixtral-8x7b (fp8),6362.0,tokens/s,LLM Inference
11557,LLM Inference: MLPerf: gptj-99 (fp8),1693.4675,tokens/s,LLM Inference
11557,LLM Inference: MLPerf: gptj-99.9 (fp8),1693.4675,tokens/s,LLM Inference
196,LLM Inference: MLPerf: gptj-99 (fp8),2414.8625,tokens/s,LLM Inference
199,LLM Inference: MLPerf: llama2-70b-99 (fp8),1222.48,tokens/s,LLM Inference
10914,LLM Inference: MLPerf: deepseek-r1 (fp4),2327.472222222222,tokens/s,LLM Inference
11557,LLM Inference: MLPerf: llama2-70b-99 (fp8),1668.9,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: distilgpt2 (bnb),62.902220141986916,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-160m (gptq),89.07948539116988,tokens/s,LLM Inference
11335,LLM Inference: MLPerf: llama3.1-8b (fp4),5757.5,tokens/s,LLM Inference
11335,LLM Inference: MLPerf: mixtral-8x7b (fp8),3766.7875,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: distilgpt2 (unquantized),181.54577762043127,tokens/s,LLM Inference
8001,LLM Inference: MLPerf: llama3.1-8b (UINT4),1337.1125,tokens/s,LLM Inference
8001,LLM Inference: MLPerf: llama2-70b-99.9 (UINT4),527.5125,tokens/s,LLM Inference
8001,LLM Inference: MLPerf: llama2-70b-99 (UINT4),527.5125,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: dummy-GPT2-correct-vocab (bnb),74.62683221969407,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neo-125m (unquantized),106.83568470792468,tokens/s,LLM Inference
195,LLM Inference: llama.cpp Bench: llama 7B Q4_0 (Q4_0),248.77,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-125m (unquantized),109.80843469283062,tokens/s,LLM Inference
11557,LLM Inference: MLPerf: mixtral-8x7b (fp8),4714.2,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt2 (gptq),108.29668832080279,tokens/s,LLM Inference
199,LLM Inference: MLPerf: mixtral-8x7b (fp8),3466.125,tokens/s,LLM Inference
11743,Hashcat: Benchmark: TrueCrypt RIPEMD160 + XTS,1281400.0,H/s,OpenBenchmarking: Hashcat
199,LLM Inference: MLPerf: gptj-99.9 (fp8),1562.8175,tokens/s,LLM Inference
199,LLM Inference: MLPerf: gptj-99 (fp8),1562.8175,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-125m (gptq),117.80638177250032,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-70m (awq),130.097533093091,tokens/s,LLM Inference
10914,LLM Inference: MLPerf: llama2-70b-99.9 (fp4),7436.4,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: distilgpt2 (awq),185.12046705502942,tokens/s,LLM Inference
195,LLM Inference: MLPerf: llama2-70b-99 (fp8),370.5775,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-125m (torchao),112.93097733128714,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-160m (unquantized),90.05382594315361,tokens/s,LLM Inference
195,LLM Inference: MLPerf: llama2-70b-99 (bfloat16),2631.93,tokens/s,LLM Inference
11743,Hashcat: Benchmark: WPA-PBKDF2-PMKID+EAPOL,1761767.0,H/s,OpenBenchmarking: Hashcat
5057,LLM Inference: LLM Perf: pythia-70m (gptq),161.3713270551849,tokens/s,LLM Inference
11335,LLM Inference: MLPerf: llama2-70b-99 (fp4),3250.7375,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: dummy-GPT2-correct-vocab (unquantized),214.42407210836075,tokens/s,LLM Inference
3625,LLM Inference: llama.cpp Bench: llama 7B Q4_0 (Q4_0),13.99,tokens/s,LLM Inference
8001,LLM Inference: MLPerf: llama3.1-8b-edge (UINT4),1652.47,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neo-125m (awq),75.20487375851783,tokens/s,LLM Inference
11335,LLM Inference: MLPerf: llama2-70b-99.9 (fp4),3250.125,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt2 (torchao),100.90186937936959,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-160m (awq),68.97287140391941,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-70m (bnb),62.66193267738452,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt2 (awq),102.00756183113688,tokens/s,LLM Inference
11557,LLM Inference: MLPerf: llama2-70b-interactive-99 (fp8),1380.1875,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-125m (awq),77.0113227367956,tokens/s,LLM Inference
10914,LLM Inference: MLPerf: llama3.1-405b (fp4),130.53,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt2-small-spanish (gptq),108.034596187476,tokens/s,LLM Inference
11199,LLM Inference: llama.cpp Bench: llama 7B Q4_0 (Q4_0),104.38,tokens/s,LLM Inference
196,LLM Inference: MLPerf: gptj-99.9 (fp8),2414.8625,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neo-125m (gptq),79.91889669400466,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt2-large (unquantized),37.93978978532382,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-2.7b (gptq),46.987471983127136,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: xglm-564M (gptq),57.99348189113929,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-350m (gptq),62.1268504210812,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: stablelm-base-alpha-7b (gptq),38.78692643031019,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: xglm-564M (awq),39.091779650814466,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt2-large (gptq),41.001026082264865,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-0.5B (unquantized),41.93668816004276,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-160m (bnb),32.76642180269336,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-1.8B (awq),30.39347887633313,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-350m (awq),41.58696743138261,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: stablelm-3b-4e1t (gptq),36.883653718338486,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: phi-1_5 (unquantized),38.35669489191812,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-2.7b (gptq),34.41986068965278,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-1.4b (gptq),45.85801893653194,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-1.8B (gptq),39.57134902941026,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: falcon-rw-1b (gptq),56.98670664622639,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: rho-math-1b-v0.1 (gptq),42.82352055579768,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-1.3b (gptq),45.695723850268834,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-0.5B (awq),30.891516688018747,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neo-1.3B (gptq),42.48871636336712,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neo-1.3B (unquantized),35.79989740242852,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neo-125m (bnb),26.416333362399538,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt2 (bnb),34.36109558269612,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-410m (unquantized),46.812698564910725,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-2.7b (awq),15.171437633290537,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neo-1.3B (awq),39.58535761479076,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: bloomz-560m (awq),40.57374375569489,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-125m (bnb),27.34458970621924,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: stablelm-base-alpha-3b (gptq),60.81624685219571,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-1.3b (unquantized),35.0798379390008,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gemma-2b (awq),31.284002711818513,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-6.7b (awq),26.259872562359202,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-410m (awq),36.69882488044026,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: xglm-564M (unquantized),58.76420749020246,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: bloomz-560m (gptq),55.4574646062635,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: stablelm-2-1_6b (gptq),45.00216676517924,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: bloomz-560m (unquantized),52.86449581573914,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-350m (unquantized),66.0573186394542,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-2.7b (awq),26.64689545669131,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: bloomz-560m (torchao),58.08748887763972,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: falcon-rw-1b (unquantized),41.66064912766921,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-1.4b (unquantized),35.097812397127626,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neo-2.7B (awq),29.95164301385483,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gemma-2b (gptq),38.3487117367116,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-6.7b (awq),6.820264567315258,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt2-small-spanish (bnb),32.93218818104824,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: phi-1_5 (gptq),42.48489621833254,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-410m (gptq),46.2650871990701,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: stablelm-2-1_6b (unquantized),35.33376359935112,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: rho-math-1b-v0.1 (unquantized),44.11929481313187,tokens/s,LLM Inference
11743,Hashcat: Benchmark: Kerberos 5 + etype 23 + TGS-REP,2110933333.0,H/s,OpenBenchmarking: Hashcat
5057,LLM Inference: LLM Perf: pythia-1.3b (awq),35.535602277664545,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-1.4b (awq),36.25599448269549,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-1.8B (unquantized),33.60188244920882,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-0.5B (gptq),41.96651319510666,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen2-0.5B (gptq),40.644603361708825,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: falcon-rw-1b (bnb),18.008522518207645,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: recurrentgemma-2b (unquantized),18.216713706343974,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: DeciLM-7B (awq),24.68891605078721,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-4B (awq),18.637961994094944,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-4B (gptq),24.759510654631082,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: xglm-4.5B (gptq),29.108557860511972,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Phi-3-mini-4k-instruct (gptq),28.538969083712495,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: xglm-7.5B (awq),24.59713901552841,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-7B (awq),21.62335189271785,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Yi-6B (gptq),31.148062257450167,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: xglm-564M (bnb),13.384654870381437,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-2.7b (unquantized),19.011681934726266,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Phi-3-mini-4k-instruct (unquantized),27.024800570506812,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-350m (bnb),13.866682142874415,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neo-1.3B (bnb),13.590899330551508,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Llama-3.1-8B-Instruct (awq),22.024580134036633,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Yi-6B (awq),23.742778833863987,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gemma-7b (gptq),25.56505399147443,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: recurrentgemma-2b (awq),17.170905533236606,tokens/s,LLM Inference
5444,Hashcat: Benchmark: LUKS v2 argon2id + SHA-256 + AES,6.0,H/s,OpenBenchmarking: Hashcat
5057,LLM Inference: LLM Perf: gemma-2b (bnb),13.72223302839092,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: llama-7b (gptq),32.329924190219174,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neo-2.7B (unquantized),18.98204390655448,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Llama-3.1-8B-Instruct (gptq),28.831747422236887,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-j-6b (awq),22.917810419774057,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-1.4b (bnb),16.74906485069438,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Mistral-7B-Instruct-v0.2 (gptq),26.41757060827416,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: stablelm-3b-4e1t (unquantized),19.78756251764159,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Meta-Llama-3-8B (gptq),28.75193068677513,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: bloomz-560m (bnb),16.715261567952655,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: xglm-7.5B (gptq),30.200307711980155,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: zephyr-7b-beta (gptq),26.996906165099148,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: xglm-4.5B (awq),20.52974961188687,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-12b (awq),18.579476050970552,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: rho-math-1b-v0.1 (bnb),12.027942705251407,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: stablelm-base-alpha-7b (bnb),12.473387684565635,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-2.7b (unquantized),19.477171159475894,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: recurrentgemma-2b (gptq),26.283213089799542,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-6.7b (gptq),33.6427975371769,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Llama-2-7b-hf (gptq),33.57341099509174,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-4B (unquantized),24.937630796737047,tokens/s,LLM Inference
11743,Hashcat: Benchmark: LUKS v2 argon2id + SHA-256 + AES,12.0,H/s,OpenBenchmarking: Hashcat
5057,LLM Inference: LLM Perf: stablelm-base-alpha-3b (unquantized),16.965786774456248,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: RedPajama-INCITE-Base-3B-v1 (unquantized),19.084676722928304,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: DeciLM-7B (gptq),26.53512600776033,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-1.3b (bnb),16.409251598216578,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Meta-Llama-3-8B-Instruct (gptq),29.14405105223489,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gemma-2b (unquantized),22.624529602530444,tokens/s,LLM Inference
5444,Hashcat: Benchmark: WPA-PBKDF2-PMKID+EAPOL,651833.0,H/s,OpenBenchmarking: Hashcat
5057,LLM Inference: LLM Perf: pythia-410m (bnb),16.682658243965452,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: stablelm-base-alpha-3b (bnb),24.137912971509092,tokens/s,LLM Inference
8929,Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024,224.52,Tokens Per Second,OpenBenchmarking: Llama.cpp
5057,LLM Inference: LLM Perf: gpt-neo-2.7B (gptq),31.835626470273276,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-6.7b (gptq),37.82486005682729,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-2.7b (bnb),10.479961082877653,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: stablelm-3b-4e1t (bnb),8.726620414757821,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: codegen-16B-nl (gptq),18.285340985917088,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gemma-2-2b (unquantized),19.43155488034343,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Yi-6B (bnb),8.351559378811086,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: phi-1_5 (bnb),12.90791142001812,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-6.7b (unquantized),15.360580776244586,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen2-beta-14B (gptq),21.633322560042426,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: xglm-4.5B (unquantized),20.21024060554282,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt2-large (bnb),11.930454614918235,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: recurrentgemma-2b (bnb),9.670418345877279,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: zephyr-7b-beta (unquantized),14.370577205487201,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neox-20b (gptq),14.523028693078363,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-13b (gptq),23.329916287036877,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: polyglot-ko-12.8b (gptq),21.392921533521495,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Mistral-7B-v0.1 (unquantized),15.458783559291824,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: stablelm-2-12b (gptq),13.178907447898812,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-14B (awq),16.928603492748614,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: stablelm-2-1_6b (bnb),11.43366358954509,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: recurrentgemma-9b (gptq),17.505536605568267,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Mistral-7B-v0.1 (gptq),25.666593862028954,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: polyglot-ko-12.8b (awq),16.957548208537624,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-0.5B (bnb),11.075410572149973,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-14B (gptq),21.565679821142446,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neo-2.7B (bnb),10.158829820971569,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-2.7b (bnb),12.421904103909288,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gemma-2-2b (gptq),20.96545715718482,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-12b (gptq),23.245383967399746,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: codegen-6B-nl (gptq),24.792188354685013,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-13b (awq),3.283837846212853,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: codegen-16B-nl (awq),16.916300557522455,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: llama-13b (gptq),23.92633971897566,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-j-6b (unquantized),17.12962483540696,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: RedPajama-INCITE-Base-3B-v1 (bnb),12.30083351818492,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Llama-2-7b-hf (unquantized),16.692342433189648,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: llama-7b (unquantized),16.71546241941083,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen2-0.5B (bnb),10.573205075639954,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Yi-6B (unquantized),18.709420616121477,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen2-beta-14B (awq),16.781797540263018,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Llama-2-13b-hf (gptq),24.18360692806807,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gemma-2-2b (torchao),19.651831278525076,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-6.7b (unquantized),16.60712745979335,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: DeciLM-7B (unquantized),16.238670293240943,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-7B (unquantized),15.696714807545138,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: recurrentgemma-9b (awq),13.388768018642995,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neox-20b (awq),12.164310510317263,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: zephyr-7b-beta (bnb),13.141799696865908,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: RedPajama-INCITE-Base-7B-v0.1 (unquantized),15.487227946322562,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Phi-3-mini-4k-instruct (bnb),10.87154862203728,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: codegen-6B-nl (unquantized),14.50501664427726,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-1.8B (bnb),11.05495808623993,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-4B (bnb),6.704430357251609,tokens/s,LLM Inference
10335,CloverLeaf: Input: clover_bm,108.7,Seconds,OpenBenchmarking: Llama.cpp
10335,CloverLeaf: Input: clover_bm16,802.16,Seconds,OpenBenchmarking: Llama.cpp
11199,LLM Inference: Geekbench AI: Geekbench AI Half Precision (fp16),54998.0,points,LLM Inference
5057,LLM Inference: LLM Perf: falcon-7b (unquantized),13.99234125800327,tokens/s,LLM Inference
11199,LLM Inference: Geekbench AI: Geekbench AI Single Precision (fp32),38235.0,points,LLM Inference
195,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),280.74,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Meta-Llama-3-8B-Instruct (bnb),7.758202776345115,tokens/s,LLM Inference
3625,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),13.99,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: codegen-6B-nl (bnb),10.518386118391225,tokens/s,LLM Inference
10913,LLM Inference: MLPerf: llama2-70b-interactive-99 (fp8),2422.625,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: xglm-7.5B (bnb),10.055417175974927,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-6.7b (bnb),12.350829729476395,tokens/s,LLM Inference
1819,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),14.27,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-MoE-A2.7B (bnb),1.7149273173013857,tokens/s,LLM Inference
10913,LLM Inference: MLPerf: llama2-70b-interactive-99.9 (fp8),1524.05,tokens/s,LLM Inference
11445,LLM Inference: Geekbench AI: Geekbench AI Half Precision (fp16),58886.0,points,LLM Inference
5057,LLM Inference: LLM Perf: DeciLM-7B (bnb),8.560504271981838,tokens/s,LLM Inference
10913,LLM Inference: MLPerf: llama2-70b-interactive-99.9 (fp16),1995.0125,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Llama-3.1-8B-Instruct (bnb),7.683537903321721,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Mistral-7B-v0.1 (bnb),8.01269058559718,tokens/s,LLM Inference
8001,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),8.21,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-neox-20b (bnb),4.6033740693976615,tokens/s,LLM Inference
10810,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),22.41,tokens/s,LLM Inference
11445,LLM Inference: Geekbench AI: Geekbench AI Single Precision (fp32),39390.0,points,LLM Inference
5057,LLM Inference: LLM Perf: gemma-7b (bnb),8.868820640420354,tokens/s,LLM Inference
3327,LLM Inference: llama.cpp Bench: qwen3 32B Q4_K - Medium (Q4_0),7.19,tokens/s,LLM Inference
10913,LLM Inference: MLPerf: llama3.1-405b (fp8),25.347,tokens/s,LLM Inference
3601,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),74.66,tokens/s,LLM Inference
11199,LLM Inference: Geekbench AI: Geekbench AI Quantized (int8),28646.0,points,LLM Inference
10913,LLM Inference: MLPerf: gptj-99 (fp8),2562.725,tokens/s,LLM Inference
11335,LLM Inference: Geekbench AI: Geekbench AI Half Precision (fp16),53322.0,points,LLM Inference
11445,LLM Inference: Geekbench AI: Geekbench AI Quantized (int8),29332.0,points,LLM Inference
5057,LLM Inference: LLM Perf: llama-7b (bnb),8.284234457318597,tokens/s,LLM Inference
11335,LLM Inference: Geekbench AI: Geekbench AI Single Precision (fp32),37299.0,points,LLM Inference
5057,LLM Inference: LLM Perf: Mistral_Pro_8B_v0.1 (bnb),6.122838711347836,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: recurrentgemma-9b (bnb),6.269754130530485,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: falcon-7b (bnb),11.29793416121512,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-6.7b (bnb),10.366066673032979,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: stablelm-2-12b (bnb),4.920548533158335,tokens/s,LLM Inference
3327,LLM Inference: llama.cpp Bench: gpt-oss 20B Q4_K - Medium (Q4_0),46.99,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen1.5-7B (bnb),8.04054537736038,tokens/s,LLM Inference
3860,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),232.92,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-j-6b (bnb),10.148739314264132,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Meta-Llama-3-8B (bnb),7.9641536545131695,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: codegen-16B-nl (bnb),5.860508452524349,tokens/s,LLM Inference
11199,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),104.38,tokens/s,LLM Inference
3327,LLM Inference: llama.cpp Bench: llama 7B Q4_0 (Q4_0),36.68,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: pythia-12b (bnb),7.797834846156245,tokens/s,LLM Inference
3327,LLM Inference: llama.cpp Bench: qwen3moe 30B.A3B Q4_K - Medium (Q4_0),35.09,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: DeepSeek-V2-Lite (bnb),9.635919683014212,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: RedPajama-INCITE-Base-7B-v0.1 (bnb),12.032295724703118,tokens/s,LLM Inference
11335,LLM Inference: Geekbench AI: Geekbench AI Quantized (int8),28264.0,points,LLM Inference
10335,"srsRAN Project: Test: PDSCH Processor Benchmark, Throughput Total",87979.2,Mbps,OpenBenchmarking: Llama.cpp
203,LLM Inference: Geekbench AI: Geekbench AI Quantized (int8),29434.0,points,LLM Inference
202,LLM Inference: Geekbench AI: Geekbench AI Quantized (int8),28261.0,points,LLM Inference
10937,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),273.68,tokens/s,LLM Inference
203,LLM Inference: Geekbench AI: Geekbench AI Half Precision (fp16),53882.0,points,LLM Inference
10910,vkpeak: fp64-scalar,813.8,GFLOPS,OpenBenchmarking: Hashcat
3327,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),41.68,tokens/s,LLM Inference
10937,LLM Inference: Geekbench AI: Geekbench AI Single Precision (fp32),46917.0,points,LLM Inference
3379,Hashcat: Benchmark: WPA-PBKDF2-PMKID+EAPOL,104967.0,H/s,OpenBenchmarking: Hashcat
5057,LLM Inference: LLM Perf: Qwen1.5-14B (bnb),7.139665918484011,tokens/s,LLM Inference
5057,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),46.38,tokens/s,LLM Inference
203,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),188.96,tokens/s,LLM Inference
203,LLM Inference: MLPerf: llama3.1-8b-edge (fp32),44.6708,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gpt-j-6b (gptq),22.20095838679643,tokens/s,LLM Inference
10910,vkpeak: fp64-vec4,812.64,GFLOPS,OpenBenchmarking: Hashcat
10910,vkpeak: int32-scalar,25964.43,GIOPS,OpenBenchmarking: Hashcat
203,LLM Inference: llama.cpp Bench: llama 7B Q4_0 (Q4_0),154.68,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gemma-2-2b (awq),17.928870059449828,tokens/s,LLM Inference
10910,vkpeak: int32-vec4,24593.47,GIOPS,OpenBenchmarking: Hashcat
10937,LLM Inference: llama.cpp Bench: llama 7B Q4_0 (Q4_0),315.3,tokens/s,LLM Inference
10910,vkpeak: int16-scalar,25955.42,GIOPS,OpenBenchmarking: Hashcat
10910,vkpeak: int16-vec4,24589.36,GIOPS,OpenBenchmarking: Hashcat
10910,vkpeak: int8-dotprod,32808.08,GIOPS,OpenBenchmarking: Hashcat
5057,LLM Inference: LLM Perf: Mistral_Pro_8B_v0.1 (gptq),22.15622060526971,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: codegen-6B-nl (awq),16.3695511428579,tokens/s,LLM Inference
10910,vkpeak: int8-matrix,409762.38,GIOPS,OpenBenchmarking: Hashcat
10910,vkpeak: bf16-matrix,101289.72,GFLOPS,OpenBenchmarking: Hashcat
10910,RealSR-NCNN: Scale: 4x - TAA: No,5.589,Seconds,OpenBenchmarking: Hashcat
10910,Waifu2x-NCNN Vulkan: Scale: 2x - Denoise: 3 - TAA: No,5.878,Seconds,OpenBenchmarking: Hashcat
10910,Hashcat: Benchmark: MD5,90224866667.0,H/s,OpenBenchmarking: Hashcat
10910,Hashcat: Benchmark: NTLM,139566666667.0,H/s,OpenBenchmarking: Hashcat
10910,Hashcat: Benchmark: SHA1,27785466667.0,H/s,OpenBenchmarking: Hashcat
10910,Hashcat: Benchmark: 7-Zip,1441767.0,H/s,OpenBenchmarking: Hashcat
202,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),191.57,tokens/s,LLM Inference
10243,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),96.65,tokens/s,LLM Inference
8827,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),47.06,tokens/s,LLM Inference
202,LLM Inference: llama.cpp Bench: llama 7B Q4_0 (Q4_0),173.16,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Mistral_Pro_8B_v0.1 (awq),14.21346978311327,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gemma-7b (awq),20.51415982074977,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: gemma-2-2b (bnb),7.900127410250198,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Llama-2-13b-hf (bnb),6.732758155908061,tokens/s,LLM Inference
10913,LLM Inference: MLPerf: llama3.1-405b (fp16),40.817125,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen2-0.5B (unquantized),37.6398130057684,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Meta-Llama-3-8B (awq),22.44153612753104,tokens/s,LLM Inference
10937,LLM Inference: Geekbench AI: Geekbench AI Half Precision (fp16),69765.0,points,LLM Inference
10937,LLM Inference: Geekbench AI: Geekbench AI Quantized (int8),34610.0,points,LLM Inference
5057,LLM Inference: LLM Perf: polyglot-ko-12.8b (bnb),7.076399080344542,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: opt-13b (bnb),7.119850900751307,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: xglm-4.5B (bnb),6.764507925459048,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: llama-13b (bnb),6.680260176003608,tokens/s,LLM Inference
5057,LLM Inference: LLM Perf: Qwen2-beta-14B (bnb),7.1268031186392,tokens/s,LLM Inference
202,LLM Inference: Geekbench AI: Geekbench AI Half Precision (fp16),53067.0,points,LLM Inference
202,LLM Inference: Geekbench AI: Geekbench AI Single Precision (fp32),37499.0,points,LLM Inference
10910,Hashcat: Benchmark: bcrypt,128000.0,H/s,OpenBenchmarking: Hashcat
10910,Hashcat: Benchmark: SHA-256,11578933333.0,H/s,OpenBenchmarking: Hashcat
10910,Hashcat: Benchmark: Kerberos 5 + etype 23 + TGS-REP,1804800000.0,H/s,OpenBenchmarking: Hashcat
10910,Rodinia: Test: OpenMP HotSpot3D,39.376,Seconds,OpenBenchmarking: Hashcat
10910,Rodinia: Test: OpenMP Leukocyte,31.639,Seconds,OpenBenchmarking: Hashcat
10910,Rodinia: Test: OpenMP CFD Solver,9.038,Seconds,OpenBenchmarking: Hashcat
10910,Rodinia: Test: OpenMP Streamcluster,11.165,Seconds,OpenBenchmarking: Hashcat
10910,Rodinia: Test: OpenCL Particle Filter,2.921,Seconds,OpenBenchmarking: Hashcat
10910,FinanceBench: Benchmark: Repo OpenMP,16807.688802,ms,OpenBenchmarking: Hashcat
10910,FinanceBench: Benchmark: Bonds OpenMP,28291.983724,ms,OpenBenchmarking: Hashcat
10910,FinanceBench: Benchmark: Monte-Carlo OpenCL,102.072001,ms,OpenBenchmarking: Hashcat
10910,FinanceBench: Benchmark: Black-Scholes OpenCL,4.174,ms,OpenBenchmarking: Hashcat
10910,ViennaCL: Test: OpenCL BLAS - sCOPY,334.0,GB/s,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: squeezenet_ssd,9.84,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: regnety_400m,23.62,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: vision_transformer,77.23,ms,OpenBenchmarking: Hashcat
10910,NCNN: Target: Vulkan GPU - Model: FastestDet,5.21,ms,OpenBenchmarking: Hashcat
10910,Blender: Blend File: Junkshop - Compute: NVIDIA CUDA,15.22,Seconds,OpenBenchmarking: Hashcat
10910,Blender: Blend File: Fishy Cat - Compute: NVIDIA CUDA,20.6,Seconds,OpenBenchmarking: Hashcat
10910,Blender: Blend File: Junkshop - Compute: NVIDIA OptiX,8.28,Seconds,OpenBenchmarking: Hashcat
10910,Blender: Blend File: Barbershop - Compute: NVIDIA CUDA,78.09,Seconds,OpenBenchmarking: Hashcat
10335,Redis: Test: LPUSH - Parallel Connections: 500,2100575.5,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Redis: Test: LPOP - Parallel Connections: 50,3768749.5,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Redis: Test: SADD - Parallel Connections: 50,3566842.75,Requests Per Second,OpenBenchmarking: Llama.cpp
3860,LLM Inference: MLPerf: llama2-70b-99 (fp8),1056.8925,tokens/s,LLM Inference
10335,Redis: Test: SET - Parallel Connections: 500,2494138.75,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Redis: Test: LPOP - Parallel Connections: 500,3765910.75,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,vLLM CPU: Test: Hermes-3-Llama-3.1-8B Throughput,121.49,tokens/s,OpenBenchmarking: Llama.cpp
10335,Redis: Test: LPUSH - Parallel Connections: 50,2125037.25,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,OpenRadioss: Model: Ford Taurus 10M,4462.67,Seconds,OpenBenchmarking: Llama.cpp
10335,Redis: Test: SADD - Parallel Connections: 500,3563792.0,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Redis: Test: SET - Parallel Connections: 50,2436053.75,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,"srsRAN Project: Test: PUSCH Processor Benchmark, Throughput Thread",348.6,Mbps,OpenBenchmarking: Llama.cpp
10335,Redis: Test: GET - Parallel Connections: 500,3167564.25,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128,29.87,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 2048,85.52,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,vLLM CPU: Test: Hermes-3-Llama-3.1-8B Latency,24.87,99p Latency - Seconds,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,1098.57,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,14.04,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: SET - Parallel Connections: 50,828267.06,Requests Per Second,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: GET - Parallel Connections: 500,1127446.62,Requests Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128,40.17,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128,22.76,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512,1419.88,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Text Generation 128,15.64,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,25.86,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024,1427.2,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,vLLM CPU: Test: Qwen2-VL-7B-Instruct Latency,22.74,99p Latency - Seconds,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048,1416.03,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024,62.47,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512,130.11,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128,70.92,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512,1363.67,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048,57.32,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Text Generation 128,85.06,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024,175.19,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512,38.15,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512,1114.99,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048,159.19,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,25.79,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128,78.56,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048,1030.45,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512,117.26,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,116.44,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048,113.97,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128,40.17,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,Valkey: Test: GET - Parallel Connections: 50,1849728.12,Requests Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024,128.17,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048,1060.54,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512,225.25,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512,1726.09,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512,1419.88,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512,90.28,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512,1138.88,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,26.17,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024,224.52,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048,1680.48,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024,88.56,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512,1129.73,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128,35.79,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,25.86,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024,1100.0,Tokens Per Second,OpenBenchmarking: Llama.cpp
9882,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512,118.8,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128,46.2,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024,1427.2,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512,1363.67,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048,1416.03,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512,130.11,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128,70.92,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024,128.17,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048,124.31,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128,37.59,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,1337.74,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128,71.45,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048,1278.18,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512,953.45,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128,19.7,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024,914.59,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048,834.67,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048,212.58,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512,1129.73,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024,1721.44,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048,1680.48,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024,88.56,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128,35.79,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048,83.29,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512,30.19,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,OpenRadioss: Model: INIVOL and Fluid Structure Interaction Drop Container,120.32,Seconds,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128,59.44,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024,1100.0,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,14.25,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024,30.14,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048,964.38,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048,28.95,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048,1060.54,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: Vulkan - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128,27.44,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512,1138.88,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512,237.79,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512,118.8,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024,1079.59,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024,120.17,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048,113.19,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,26.29,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024,236.95,Tokens Per Second,OpenBenchmarking: Llama.cpp
8929,Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048,230.31,Tokens Per Second,OpenBenchmarking: Llama.cpp
10335,"srsRAN Project: Test: PDSCH Processor Benchmark, Throughput Thread",2442.4,Mbps,OpenBenchmarking: Llama.cpp
10335,"srsRAN Project: Test: PUSCH Processor Benchmark, Throughput Total",13400.0,Mbps,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512,44.18,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,57.35,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024,42.81,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048,39.72,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512,12.7,Tokens Per Second,OpenBenchmarking: Llama.cpp
3864,LLM Inference: llama.cpp Bench: llama-2-7b-Q4_0 (Q4_0),63.94,tokens/s,LLM Inference
11557,LLM Inference: MLPerf: llama2-70b-interactive-99.9 (fp8),1380.1875,tokens/s,LLM Inference
208,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,5.58,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024,12.6,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048,12.3,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512,4190.62,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024,4188.87,Tokens Per Second,OpenBenchmarking: Llama.cpp
5057,LLM Inference: LLM Perf: Llama-2-7b-hf (bnb),8.54757352298474,tokens/s,LLM Inference
208,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,56.08,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,5.49,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512,3892.24,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,3882.77,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048,3828.48,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128,155.05,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512,41.51,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,39.94,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048,38.71,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128,16.81,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,54.27,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512,53.59,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512,4961.29,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024,53.29,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048,51.48,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,4921.01,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048,4826.78,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128,16.93,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128,7.54,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048,4136.36,Tokens Per Second,OpenBenchmarking: Llama.cpp
10914,LLM Inference: MLPerf: llama2-70b-99 (fp4),7436.4,tokens/s,LLM Inference
203,LLM Inference: Geekbench AI: Geekbench AI Single Precision (fp32),39357.0,points,LLM Inference
208,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512,41.98,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024,41.6,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048,39.61,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,55.49,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128,59.78,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128,5.86,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512,5119.58,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 512,4235.73,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024,5116.51,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048,5021.07,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 1024,4214.36,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 2048,4151.86,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 512,41.81,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Text Generation 128,57.35,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 1024,41.38,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 2048,39.64,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128,58.41,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Text Generation 128,326.99,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Text Generation 128,5.57,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Prompt Processing 512,4186.21,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Text Generation 128,43.07,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 512,5184.92,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Prompt Processing 1024,4190.6,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Prompt Processing 2048,4137.76,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 512,13322.05,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Prompt Processing 512,42.05,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 1024,5130.36,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 2048,5038.34,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 1024,13215.02,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: Vulkan - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 2048,13017.55,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Prompt Processing 1024,41.26,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Prompt Processing 2048,39.72,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 512,142.02,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Text Generation 128,55.49,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 1024,142.82,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 2048,134.6,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Text Generation 128,267.26,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Prompt Processing 512,5144.11,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Prompt Processing 1024,5118.21,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: DeepSeek-R1-Distill-Llama-8B-Q8_0 - Test: Prompt Processing 2048,5026.94,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 512,12912.08,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 1024,12812.21,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,Llama.cpp: Backend: NVIDIA CUDA - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 2048,12535.0,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,25.79,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512,117.26,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,14.04,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512,1114.99,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512,1419.88,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,1098.57,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,116.44,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048,1030.45,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128,78.56,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048,113.97,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128,40.17,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,25.86,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048,1416.03,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024,1427.2,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512,130.11,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128,70.92,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 512,1363.67,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024,128.17,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048,124.31,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128,46.2,Tokens Per Second,OpenBenchmarking: Llama.cpp
10819,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,70.01,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,1337.74,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512,90.28,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 2048,1278.18,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128,71.45,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024,30.14,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128,37.59,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,26.17,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512,953.45,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128,19.7,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048,83.29,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024,914.59,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024,224.52,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048,834.67,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512,225.25,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 512,1726.09,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048,212.58,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512,30.19,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024,1721.44,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 2048,1680.48,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128,35.79,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024,88.56,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128,59.44,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048,28.95,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,14.25,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512,1129.73,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048,1060.54,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: Vulkan - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024,1100.0,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 512,1138.88,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512,118.8,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024,1079.59,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 2048,964.38,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 512,237.79,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024,120.17,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048,113.19,Tokens Per Second,OpenBenchmarking: Llama.cpp
10939,LLM Inference: llama.cpp Bench: llama 7B Q4_0 (Q4_0),315.85,tokens/s,LLM Inference
10683,Llama.cpp: Backend: Vulkan - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128,27.44,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,26.29,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024,236.95,Tokens Per Second,OpenBenchmarking: Llama.cpp
10683,Llama.cpp: Backend: AMD ROCm HIP - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 2048,230.31,Tokens Per Second,OpenBenchmarking: Llama.cpp
208,vkpeak: fp32-scalar,28399.8,GFLOPS,OpenBenchmarking: Hashcat
208,vkpeak: fp32-vec4,28089.05,GFLOPS,OpenBenchmarking: Hashcat
208,vkpeak: fp16-scalar,42450.18,GFLOPS,OpenBenchmarking: Hashcat
208,vkpeak: fp16-vec4,30825.53,GFLOPS,OpenBenchmarking: Hashcat
208,vkpeak: fp16-matrix,169095.91,GFLOPS,OpenBenchmarking: Hashcat
208,vkpeak: fp64-scalar,667.09,GFLOPS,OpenBenchmarking: Hashcat
208,vkpeak: fp64-vec4,666.29,GFLOPS,OpenBenchmarking: Hashcat
208,vkpeak: int32-scalar,21162.25,GIOPS,OpenBenchmarking: Hashcat
208,vkpeak: int32-vec4,21113.67,GIOPS,OpenBenchmarking: Hashcat
208,vkpeak: int16-scalar,18057.6,GIOPS,OpenBenchmarking: Hashcat
208,vkpeak: int16-vec4,16143.7,GIOPS,OpenBenchmarking: Hashcat
208,vkpeak: int8-dotprod,19992.46,GIOPS,OpenBenchmarking: Hashcat
208,vkpeak: int8-matrix,335774.83,GIOPS,OpenBenchmarking: Hashcat
208,vkpeak: bf16-matrix,83386.17,GFLOPS,OpenBenchmarking: Hashcat
208,RealSR-NCNN: Scale: 4x - TAA: No,6.348,Seconds,OpenBenchmarking: Hashcat
208,RealSR-NCNN: Scale: 4x - TAA: Yes,33.828,Seconds,OpenBenchmarking: Hashcat
208,Waifu2x-NCNN Vulkan: Scale: 2x - Denoise: 3 - TAA: No,5.965,Seconds,OpenBenchmarking: Hashcat
208,Waifu2x-NCNN Vulkan: Scale: 4x - Denoise: 3 - TAA: No,20.694,Seconds,OpenBenchmarking: Hashcat
208,Waifu2x-NCNN Vulkan: Scale: 2x - Denoise: 3 - TAA: Yes,12.654,Seconds,OpenBenchmarking: Hashcat
208,Waifu2x-NCNN Vulkan: Scale: 4x - Denoise: 3 - TAA: Yes,54.323,Seconds,OpenBenchmarking: Hashcat
208,"NAMD CUDA: Input: ATPase with 327,506 Atoms",13.04681,ns/day,OpenBenchmarking: Hashcat
208,"NAMD CUDA: Input: STMV with 1,066,628 Atoms",3.50394,ns/day,OpenBenchmarking: Hashcat
5057,LLM Inference: LLM Perf: gpt2-small-spanish (unquantized),103.80659219711217,tokens/s,LLM Inference
208,OctaneBench: Total Score,697.96351,Score,OpenBenchmarking: Hashcat
208,Blender: Blend File: BMW27 - Compute: NVIDIA OptiX,5.52,Seconds,OpenBenchmarking: Hashcat
208,Blender: Blend File: Junkshop - Compute: NVIDIA OptiX,8.46,Seconds,OpenBenchmarking: Hashcat
208,Blender: Blend File: Classroom - Compute: NVIDIA OptiX,11.89,Seconds,OpenBenchmarking: Hashcat
208,Blender: Blend File: Fishy Cat - Compute: NVIDIA OptiX,9.51,Seconds,OpenBenchmarking: Hashcat
208,Blender: Blend File: Barbershop - Compute: NVIDIA OptiX,54.39,Seconds,OpenBenchmarking: Hashcat
208,Blender: Blend File: Pabellon Barcelona - Compute: NVIDIA OptiX,13.75,Seconds,OpenBenchmarking: Hashcat
3667,Hashcat: Benchmark: SHA-256,1219833333.0,H/s,OpenBenchmarking: Hashcat
11889,vkpeak: fp32-scalar,34789.41,GFLOPS,OpenBenchmarking: Hashcat
11889,Blender: Blend File: Junkshop - Compute: NVIDIA CUDA,15.22,Seconds,OpenBenchmarking: Hashcat
11743,GravityMark: Resolution: 6144 x 2560 - Renderer: OpenGL,175.6,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,GravityMark: Resolution: 6144 x 2560 - Renderer: Vulkan,188.1,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,GravityMark: Resolution: 6144 x 2560 - Renderer: Vulkan Ray-Tracing,154.6,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,3DMark Wild Life Extreme: Resolution: 6144 x 2560,188.96,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,Unigine Superposition: Resolution: 6144 x 2560 - Mode: Fullscreen - Quality: Low - Renderer: OpenGL,194.4,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,Unigine Superposition: Resolution: 6144 x 2560 - Mode: Fullscreen - Quality: Medium - Renderer: OpenGL,81.7,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,Unvanquished: Resolution: 6144 x 2560 - Effects Quality: High,473.5,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,Unvanquished: Resolution: 6144 x 2560 - Effects Quality: Ultra,502.8,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,Unvanquished: Resolution: 6144 x 2560 - Effects Quality: Medium,490.2,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,Xonotic: Resolution: 6144 x 2560 - Effects Quality: Low,702.6719706,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,Xonotic: Resolution: 6144 x 2560 - Effects Quality: High,310.7002068,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,Xonotic: Resolution: 6144 x 2560 - Effects Quality: Ultra,261.292326,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,Xonotic: Resolution: 6144 x 2560 - Effects Quality: Ultimate,208.4845102,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,SVT-AV1: Encoder Mode: Preset 3 - Input: Bosphorus 4K,21.975,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,SVT-AV1: Encoder Mode: Preset 5 - Input: Bosphorus 4K,57.43,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,SVT-AV1: Encoder Mode: Preset 8 - Input: Bosphorus 4K,110.678,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,SVT-AV1: Encoder Mode: Preset 13 - Input: Bosphorus 4K,179.688,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,SVT-AV1: Encoder Mode: Preset 3 - Input: Beauty 4K 10-bit,3.953,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,SVT-AV1: Encoder Mode: Preset 5 - Input: Beauty 4K 10-bit,11.762,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,SVT-AV1: Encoder Mode: Preset 8 - Input: Beauty 4K 10-bit,17.737,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,SVT-AV1: Encoder Mode: Preset 13 - Input: Beauty 4K 10-bit,21.281,Frames Per Second,OpenBenchmarking: Llama.cpp
11743,SVT-JPEG-XS: Input: Bosphorus 4K,11.01,Latency (ms),OpenBenchmarking: Llama.cpp
11743,SVT-JPEG-XS: Input: Beauty 4K 10-bit,2.99,Latency (ms),OpenBenchmarking: Llama.cpp
11743,Timed Godot Game Engine Compilation: Time To Compile,205.105,Seconds,OpenBenchmarking: Llama.cpp
11743,Timed Linux Kernel Compilation: Build: defconfig,50.706,Seconds,OpenBenchmarking: Llama.cpp
11743,Timed Eigen Compilation: Time To Compile,43.349,Seconds,OpenBenchmarking: Llama.cpp
11743,Timed Erlang/OTP Compilation: Time To Compile,19.029,Seconds,OpenBenchmarking: Llama.cpp
11743,Node.js V8 Web Tooling Benchmark,29.94,runs/s,OpenBenchmarking: Llama.cpp
11743,"ClickHouse: 100M Rows Hits Dataset, Latency, Second Run",480.96,"Queries Per Minute, Geo Mean",OpenBenchmarking: Llama.cpp
11743,"ClickHouse: 100M Rows Hits Dataset, Latency, Third Run",483.0,"Queries Per Minute, Geo Mean",OpenBenchmarking: Llama.cpp
11743,"ClickHouse: 100M Rows Hits Dataset, Throughput",97.28,"Queries Per Minute, Geo Mean",OpenBenchmarking: Llama.cpp
11743,Apache CouchDB: Bulk Size: 100 - Inserts: 1000 - Rounds: 30,48.733,Seconds,OpenBenchmarking: Llama.cpp
11743,Apache CouchDB: Bulk Size: 100 - Inserts: 3000 - Rounds: 30,150.943,Seconds,OpenBenchmarking: Llama.cpp
11743,"CockroachDB: Workload: KV, 60% Reads - Concurrency: 128",99295.5,ops/s,OpenBenchmarking: Llama.cpp
11743,"CockroachDB: Workload: KV, 60% Reads - Concurrency: 256",103972.9,ops/s,OpenBenchmarking: Llama.cpp
11743,"CockroachDB: Workload: KV, 95% Reads - Concurrency: 128",160751.3,ops/s,OpenBenchmarking: Llama.cpp
11743,"CockroachDB: Workload: KV, 95% Reads - Concurrency: 256",157019.2,ops/s,OpenBenchmarking: Llama.cpp
11743,TigerBeetle: Clients: 1,117.0,ms - query latency p99,OpenBenchmarking: Llama.cpp
11743,TigerBeetle: Clients: 4,501.0,ms - query latency p99,OpenBenchmarking: Llama.cpp
11743,TigerBeetle: Clients: 8,939.0,ms - query latency p99,OpenBenchmarking: Llama.cpp
11743,PostgreSQL: Scaling Factor: 1 - Clients: 800 - Mode: Read Only,1963379.0,TPS,OpenBenchmarking: Llama.cpp
11743,PostgreSQL: Scaling Factor: 1 - Clients: 800 - Mode: Read Only - Average Latency,0.408,ms,OpenBenchmarking: Llama.cpp
11743,PostgreSQL: Scaling Factor: 1 - Clients: 800 - Mode: Read Write,982.0,TPS,OpenBenchmarking: Llama.cpp
11743,PostgreSQL: Scaling Factor: 1 - Clients: 800 - Mode: Read Write - Average Latency,814.998,ms,OpenBenchmarking: Llama.cpp
11743,PostgreSQL: Scaling Factor: 100 - Clients: 800 - Mode: Read Only,1702959.0,TPS,OpenBenchmarking: Llama.cpp
11743,PostgreSQL: Scaling Factor: 100 - Clients: 800 - Mode: Read Only - Average Latency,0.47,ms,OpenBenchmarking: Llama.cpp
11743,PostgreSQL: Scaling Factor: 100 - Clients: 800 - Mode: Read Write,41192.0,TPS,OpenBenchmarking: Llama.cpp
11743,PostgreSQL: Scaling Factor: 100 - Clients: 800 - Mode: Read Write - Average Latency,19.421,ms,OpenBenchmarking: Llama.cpp
11743,GNU Octave Benchmark,4.847,Seconds,OpenBenchmarking: Llama.cpp
11743,RawTherapee: Total Benchmark Time,42.777,Seconds,OpenBenchmarking: Llama.cpp
11743,Blender: Blend File: Junkshop - Compute: CPU-Only,60.5,Seconds,OpenBenchmarking: Llama.cpp
11743,Blender: Blend File: Junkshop - Compute: NVIDIA CUDA,14.43,Seconds,OpenBenchmarking: Llama.cpp
11743,Blender: Blend File: Junkshop - Compute: NVIDIA OptiX,7.33,Seconds,OpenBenchmarking: Llama.cpp
11743,PyBench: Total For Average Test Times,335.0,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: go,46.7,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: chaos,22.4,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: float,27.1,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: nbody,36.5,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: pathlib,5.95,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: raytrace,103.0,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: xml_etree,22.4,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: gc_collect,867.0,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: json_loads,9.53,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: crypto_pyaes,30.2,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: regex_compile,48.7,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: python_startup,5.12,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: asyncio_tcp_ssl,413.0,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: django_template,14.5,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: asyncio_websockets,273.0,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PyPerformance: Benchmark: pickle_pure_python,123.0,Milliseconds,OpenBenchmarking: Llama.cpp
11743,PHPBench: PHP Benchmark Suite,1570764.0,Score,OpenBenchmarking: Llama.cpp
11743,Selenium: Benchmark: StyleBench - Browser: Firefox,206.0,Runs / Minute,OpenBenchmarking: Llama.cpp
11743,Selenium: Benchmark: Jetstream 2 - Browser: Firefox,365.255,Score,OpenBenchmarking: Llama.cpp
11743,Selenium: Benchmark: PSPDFKit WASM - Browser: Firefox,249.0,Score,OpenBenchmarking: Llama.cpp
11743,Selenium: Benchmark: Speedometer 3.1 - Browser: Firefox,38.0,Score,OpenBenchmarking: Llama.cpp
11743,Selenium: Benchmark: WASM imageConvolute - Browser: Firefox,15.9,ms,OpenBenchmarking: Llama.cpp
11743,Selenium: Benchmark: WASM collisionDetection - Browser: Firefox,213.3,ms,OpenBenchmarking: Llama.cpp
11743,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Text Generation 128,5.69,Tokens Per Second,OpenBenchmarking: Llama.cpp
11743,Llama.cpp: Backend: CPU BLAS - Model: Qwen3-8B-Q8_0 - Test: Prompt Processing 1024,86.73,Tokens Per Second,OpenBenchmarking: Llama.cpp
11743,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Text Generation 128,17.18,Tokens Per Second,OpenBenchmarking: Llama.cpp
11743,Llama.cpp: Backend: CPU BLAS - Model: gpt-oss-20b-Q8_0 - Test: Prompt Processing 1024,141.76,Tokens Per Second,OpenBenchmarking: Llama.cpp
11743,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Text Generation 128,18.27,Tokens Per Second,OpenBenchmarking: Llama.cpp
11743,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Text Generation 128,11.3,Tokens Per Second,OpenBenchmarking: Llama.cpp
11743,Llama.cpp: Backend: CPU BLAS - Model: GLM-4.7-Flash-IQ4_XS - Test: Prompt Processing 1024,88.62,Tokens Per Second,OpenBenchmarking: Llama.cpp
11743,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128,5.76,Tokens Per Second,OpenBenchmarking: Llama.cpp
11743,Llama.cpp: Backend: CPU BLAS - Model: MiniMax-M2.5-UD-TQ1_0 - Test: Prompt Processing 1024,32.56,Tokens Per Second,OpenBenchmarking: Llama.cpp
11743,Llama.cpp: Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024,90.77,Tokens Per Second,OpenBenchmarking: Llama.cpp
11743,Llama.cpp: Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128,6.07,Tokens Per Second,OpenBenchmarking: Llama.cpp
11743,Llama.cpp: Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 1024,91.08,Tokens Per Second,OpenBenchmarking: Llama.cpp