code
large_stringlengths
193
899
workload_type
large_stringclasses
15 values
gpu_name
large_stringclasses
12 values
gpu_features
large_stringclasses
12 values
runtime_ms
float64
0.01
4.35k
flops
float64
30
21,045B
memory_bytes
int64
160
13.2B
arithmetic_intensity
float64
0
6.37k
dtype_bytes
int64
2
4
workload_params
large_stringlengths
11
79
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 2048) x (2048, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 2048, dtype=torch.float16, device='cuda') B = torch.randn(2048, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA RTX 4090
{"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72}
0.06618
2,147,483,648
18,350,080
117.028571
2
{"M": 128, "N": 4096, "K": 2048, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 2048) x (2048, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 2048, dtype=torch.float16, device='cuda') B = torch.randn(2048, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA H100 SXM
{"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50}
0.04731
2,147,483,648
18,350,080
117.028571
2
{"M": 128, "N": 4096, "K": 2048, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 2048) x (2048, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 2048, dtype=torch.float16, device='cuda') B = torch.randn(2048, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA H100 PCIe
{"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50}
0.031247
2,147,483,648
18,350,080
117.028571
2
{"M": 128, "N": 4096, "K": 2048, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 2048) x (2048, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 2048, dtype=torch.float16, device='cuda') B = torch.randn(2048, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA RTX A6000
{"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6}
0.057025
2,147,483,648
18,350,080
117.028571
2
{"M": 128, "N": 4096, "K": 2048, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float32, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA T4
{"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4}
0.894325
4,294,967,296
71,303,168
60.235294
4
{"M": 128, "N": 4096, "K": 4096, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float32, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA V100
{"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6}
0.467365
4,294,967,296
71,303,168
60.235294
4
{"M": 128, "N": 4096, "K": 4096, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float32, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA A10G
{"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6}
0.325985
4,294,967,296
71,303,168
60.235294
4
{"M": 128, "N": 4096, "K": 4096, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float32, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA A100 40GB
{"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.33981
4,294,967,296
71,303,168
60.235294
4
{"M": 128, "N": 4096, "K": 4096, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float32, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA A100 80GB
{"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.347715
4,294,967,296
71,303,168
60.235294
4
{"M": 128, "N": 4096, "K": 4096, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float32, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA L4
{"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48}
0.420115
4,294,967,296
71,303,168
60.235294
4
{"M": 128, "N": 4096, "K": 4096, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float32, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA L40S
{"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96}
0.122677
4,294,967,296
71,303,168
60.235294
4
{"M": 128, "N": 4096, "K": 4096, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float32, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA RTX 3090
{"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6}
0.208514
4,294,967,296
71,303,168
60.235294
4
{"M": 128, "N": 4096, "K": 4096, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float32, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA RTX 4090
{"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72}
0.163351
4,294,967,296
71,303,168
60.235294
4
{"M": 128, "N": 4096, "K": 4096, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float32, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA H100 SXM
{"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50}
0.150451
4,294,967,296
71,303,168
60.235294
4
{"M": 128, "N": 4096, "K": 4096, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float32, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA H100 PCIe
{"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50}
0.156265
4,294,967,296
71,303,168
60.235294
4
{"M": 128, "N": 4096, "K": 4096, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float32, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA RTX A6000
{"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6}
0.170053
4,294,967,296
71,303,168
60.235294
4
{"M": 128, "N": 4096, "K": 4096, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float16, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA T4
{"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4}
0.215676
4,294,967,296
35,651,584
120.470588
2
{"M": 128, "N": 4096, "K": 4096, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float16, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA V100
{"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6}
0.087811
4,294,967,296
35,651,584
120.470588
2
{"M": 128, "N": 4096, "K": 4096, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float16, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA A10G
{"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6}
0.128394
4,294,967,296
35,651,584
120.470588
2
{"M": 128, "N": 4096, "K": 4096, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float16, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA A100 40GB
{"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.065354
4,294,967,296
35,651,584
120.470588
2
{"M": 128, "N": 4096, "K": 4096, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float16, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA A100 80GB
{"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.074966
4,294,967,296
35,651,584
120.470588
2
{"M": 128, "N": 4096, "K": 4096, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float16, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA L4
{"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48}
0.173615
4,294,967,296
35,651,584
120.470588
2
{"M": 128, "N": 4096, "K": 4096, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float16, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA L40S
{"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96}
0.079605
4,294,967,296
35,651,584
120.470588
2
{"M": 128, "N": 4096, "K": 4096, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float16, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA RTX 3090
{"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6}
0.106997
4,294,967,296
35,651,584
120.470588
2
{"M": 128, "N": 4096, "K": 4096, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float16, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA RTX 4090
{"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72}
0.049501
4,294,967,296
35,651,584
120.470588
2
{"M": 128, "N": 4096, "K": 4096, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float16, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA H100 SXM
{"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50}
0.034879
4,294,967,296
35,651,584
120.470588
2
{"M": 128, "N": 4096, "K": 4096, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float16, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA H100 PCIe
{"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50}
0.038418
4,294,967,296
35,651,584
120.470588
2
{"M": 128, "N": 4096, "K": 4096, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (128, 4096) x (4096, 4096) -> (128, 4096) C = torch.matmul(A, B) return C A = torch.randn(128, 4096, dtype=torch.float16, device='cuda') B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchro...
matmul
NVIDIA RTX A6000
{"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6}
0.138438
4,294,967,296
35,651,584
120.470588
2
{"M": 128, "N": 4096, "K": 4096, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float32, device='cuda') B = torch.randn(64, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA T4
{"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4}
0.056619
2,097,152
147,456
14.222222
4
{"M": 256, "N": 64, "K": 64, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float32, device='cuda') B = torch.randn(64, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA V100
{"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6}
0.039894
2,097,152
147,456
14.222222
4
{"M": 256, "N": 64, "K": 64, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float32, device='cuda') B = torch.randn(64, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A10G
{"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6}
0.068227
2,097,152
147,456
14.222222
4
{"M": 256, "N": 64, "K": 64, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float32, device='cuda') B = torch.randn(64, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A100 40GB
{"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.050467
2,097,152
147,456
14.222222
4
{"M": 256, "N": 64, "K": 64, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float32, device='cuda') B = torch.randn(64, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A100 80GB
{"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.053403
2,097,152
147,456
14.222222
4
{"M": 256, "N": 64, "K": 64, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float32, device='cuda') B = torch.randn(64, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA L4
{"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48}
0.025951
2,097,152
147,456
14.222222
4
{"M": 256, "N": 64, "K": 64, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float32, device='cuda') B = torch.randn(64, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA L40S
{"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96}
0.034351
2,097,152
147,456
14.222222
4
{"M": 256, "N": 64, "K": 64, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float32, device='cuda') B = torch.randn(64, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX 3090
{"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6}
0.031648
2,097,152
147,456
14.222222
4
{"M": 256, "N": 64, "K": 64, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float32, device='cuda') B = torch.randn(64, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX 4090
{"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72}
0.053935
2,097,152
147,456
14.222222
4
{"M": 256, "N": 64, "K": 64, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float32, device='cuda') B = torch.randn(64, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA H100 SXM
{"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50}
0.033004
2,097,152
147,456
14.222222
4
{"M": 256, "N": 64, "K": 64, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float32, device='cuda') B = torch.randn(64, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA H100 PCIe
{"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50}
0.036379
2,097,152
147,456
14.222222
4
{"M": 256, "N": 64, "K": 64, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float32, device='cuda') B = torch.randn(64, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX A6000
{"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6}
0.037156
2,097,152
147,456
14.222222
4
{"M": 256, "N": 64, "K": 64, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float16, device='cuda') B = torch.randn(64, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA T4
{"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4}
0.026762
2,097,152
73,728
28.444444
2
{"M": 256, "N": 64, "K": 64, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float16, device='cuda') B = torch.randn(64, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA V100
{"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6}
0.054733
2,097,152
73,728
28.444444
2
{"M": 256, "N": 64, "K": 64, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float16, device='cuda') B = torch.randn(64, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A10G
{"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6}
0.02365
2,097,152
73,728
28.444444
2
{"M": 256, "N": 64, "K": 64, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float16, device='cuda') B = torch.randn(64, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A100 40GB
{"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.031697
2,097,152
73,728
28.444444
2
{"M": 256, "N": 64, "K": 64, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float16, device='cuda') B = torch.randn(64, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A100 80GB
{"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.058106
2,097,152
73,728
28.444444
2
{"M": 256, "N": 64, "K": 64, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float16, device='cuda') B = torch.randn(64, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA L4
{"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48}
0.040478
2,097,152
73,728
28.444444
2
{"M": 256, "N": 64, "K": 64, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float16, device='cuda') B = torch.randn(64, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA L40S
{"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96}
0.055354
2,097,152
73,728
28.444444
2
{"M": 256, "N": 64, "K": 64, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float16, device='cuda') B = torch.randn(64, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX 3090
{"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6}
0.051885
2,097,152
73,728
28.444444
2
{"M": 256, "N": 64, "K": 64, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float16, device='cuda') B = torch.randn(64, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX 4090
{"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72}
0.049796
2,097,152
73,728
28.444444
2
{"M": 256, "N": 64, "K": 64, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float16, device='cuda') B = torch.randn(64, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA H100 SXM
{"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50}
0.042488
2,097,152
73,728
28.444444
2
{"M": 256, "N": 64, "K": 64, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float16, device='cuda') B = torch.randn(64, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA H100 PCIe
{"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50}
0.052498
2,097,152
73,728
28.444444
2
{"M": 256, "N": 64, "K": 64, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 64) x (64, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 64, dtype=torch.float16, device='cuda') B = torch.randn(64, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX A6000
{"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6}
0.039677
2,097,152
73,728
28.444444
2
{"M": 256, "N": 64, "K": 64, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float32, device='cuda') B = torch.randn(256, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA T4
{"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4}
0.044478
8,388,608
393,216
21.333333
4
{"M": 256, "N": 64, "K": 256, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float32, device='cuda') B = torch.randn(256, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA V100
{"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6}
0.050395
8,388,608
393,216
21.333333
4
{"M": 256, "N": 64, "K": 256, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float32, device='cuda') B = torch.randn(256, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A10G
{"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6}
0.042645
8,388,608
393,216
21.333333
4
{"M": 256, "N": 64, "K": 256, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float32, device='cuda') B = torch.randn(256, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A100 40GB
{"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.039283
8,388,608
393,216
21.333333
4
{"M": 256, "N": 64, "K": 256, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float32, device='cuda') B = torch.randn(256, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A100 80GB
{"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.05257
8,388,608
393,216
21.333333
4
{"M": 256, "N": 64, "K": 256, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float32, device='cuda') B = torch.randn(256, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA L4
{"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48}
0.029448
8,388,608
393,216
21.333333
4
{"M": 256, "N": 64, "K": 256, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float32, device='cuda') B = torch.randn(256, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA L40S
{"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96}
0.050024
8,388,608
393,216
21.333333
4
{"M": 256, "N": 64, "K": 256, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float32, device='cuda') B = torch.randn(256, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX 3090
{"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6}
0.040305
8,388,608
393,216
21.333333
4
{"M": 256, "N": 64, "K": 256, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float32, device='cuda') B = torch.randn(256, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX 4090
{"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72}
0.027603
8,388,608
393,216
21.333333
4
{"M": 256, "N": 64, "K": 256, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float32, device='cuda') B = torch.randn(256, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA H100 SXM
{"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50}
0.04764
8,388,608
393,216
21.333333
4
{"M": 256, "N": 64, "K": 256, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float32, device='cuda') B = torch.randn(256, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA H100 PCIe
{"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50}
0.071175
8,388,608
393,216
21.333333
4
{"M": 256, "N": 64, "K": 256, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float32, device='cuda') B = torch.randn(256, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX A6000
{"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6}
0.054004
8,388,608
393,216
21.333333
4
{"M": 256, "N": 64, "K": 256, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float16, device='cuda') B = torch.randn(256, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA T4
{"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4}
0.050577
8,388,608
196,608
42.666667
2
{"M": 256, "N": 64, "K": 256, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float16, device='cuda') B = torch.randn(256, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA V100
{"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6}
0.039274
8,388,608
196,608
42.666667
2
{"M": 256, "N": 64, "K": 256, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float16, device='cuda') B = torch.randn(256, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A10G
{"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6}
0.04222
8,388,608
196,608
42.666667
2
{"M": 256, "N": 64, "K": 256, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float16, device='cuda') B = torch.randn(256, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A100 40GB
{"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.050011
8,388,608
196,608
42.666667
2
{"M": 256, "N": 64, "K": 256, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float16, device='cuda') B = torch.randn(256, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A100 80GB
{"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.038268
8,388,608
196,608
42.666667
2
{"M": 256, "N": 64, "K": 256, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float16, device='cuda') B = torch.randn(256, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA L4
{"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48}
0.053216
8,388,608
196,608
42.666667
2
{"M": 256, "N": 64, "K": 256, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float16, device='cuda') B = torch.randn(256, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA L40S
{"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96}
0.044377
8,388,608
196,608
42.666667
2
{"M": 256, "N": 64, "K": 256, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float16, device='cuda') B = torch.randn(256, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX 3090
{"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6}
0.044288
8,388,608
196,608
42.666667
2
{"M": 256, "N": 64, "K": 256, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float16, device='cuda') B = torch.randn(256, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX 4090
{"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72}
0.054154
8,388,608
196,608
42.666667
2
{"M": 256, "N": 64, "K": 256, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float16, device='cuda') B = torch.randn(256, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA H100 SXM
{"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50}
0.045521
8,388,608
196,608
42.666667
2
{"M": 256, "N": 64, "K": 256, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float16, device='cuda') B = torch.randn(256, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA H100 PCIe
{"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50}
0.047444
8,388,608
196,608
42.666667
2
{"M": 256, "N": 64, "K": 256, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 256) x (256, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 256, dtype=torch.float16, device='cuda') B = torch.randn(256, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX A6000
{"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6}
0.041572
8,388,608
196,608
42.666667
2
{"M": 256, "N": 64, "K": 256, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float32, device='cuda') B = torch.randn(512, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA T4
{"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4}
0.033946
16,777,216
720,896
23.272727
4
{"M": 256, "N": 64, "K": 512, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float32, device='cuda') B = torch.randn(512, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA V100
{"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6}
0.043904
16,777,216
720,896
23.272727
4
{"M": 256, "N": 64, "K": 512, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float32, device='cuda') B = torch.randn(512, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A10G
{"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6}
0.048386
16,777,216
720,896
23.272727
4
{"M": 256, "N": 64, "K": 512, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float32, device='cuda') B = torch.randn(512, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A100 40GB
{"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.033418
16,777,216
720,896
23.272727
4
{"M": 256, "N": 64, "K": 512, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float32, device='cuda') B = torch.randn(512, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A100 80GB
{"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.05222
16,777,216
720,896
23.272727
4
{"M": 256, "N": 64, "K": 512, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float32, device='cuda') B = torch.randn(512, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA L4
{"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48}
0.033595
16,777,216
720,896
23.272727
4
{"M": 256, "N": 64, "K": 512, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float32, device='cuda') B = torch.randn(512, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA L40S
{"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96}
0.030087
16,777,216
720,896
23.272727
4
{"M": 256, "N": 64, "K": 512, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float32, device='cuda') B = torch.randn(512, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX 3090
{"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6}
0.038605
16,777,216
720,896
23.272727
4
{"M": 256, "N": 64, "K": 512, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float32, device='cuda') B = torch.randn(512, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX 4090
{"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72}
0.029548
16,777,216
720,896
23.272727
4
{"M": 256, "N": 64, "K": 512, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float32, device='cuda') B = torch.randn(512, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA H100 SXM
{"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50}
0.046528
16,777,216
720,896
23.272727
4
{"M": 256, "N": 64, "K": 512, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float32, device='cuda') B = torch.randn(512, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA H100 PCIe
{"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50}
0.035792
16,777,216
720,896
23.272727
4
{"M": 256, "N": 64, "K": 512, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float32, device='cuda') B = torch.randn(512, 64, dtype=torch.float32, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX A6000
{"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6}
0.05834
16,777,216
720,896
23.272727
4
{"M": 256, "N": 64, "K": 512, "dtype": "float32"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float16, device='cuda') B = torch.randn(512, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA T4
{"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4}
0.027198
16,777,216
360,448
46.545455
2
{"M": 256, "N": 64, "K": 512, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float16, device='cuda') B = torch.randn(512, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA V100
{"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6}
0.05814
16,777,216
360,448
46.545455
2
{"M": 256, "N": 64, "K": 512, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float16, device='cuda') B = torch.randn(512, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A10G
{"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6}
0.045733
16,777,216
360,448
46.545455
2
{"M": 256, "N": 64, "K": 512, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float16, device='cuda') B = torch.randn(512, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A100 40GB
{"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.038005
16,777,216
360,448
46.545455
2
{"M": 256, "N": 64, "K": 512, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float16, device='cuda') B = torch.randn(512, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA A100 80GB
{"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40}
0.06619
16,777,216
360,448
46.545455
2
{"M": 256, "N": 64, "K": 512, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float16, device='cuda') B = torch.randn(512, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA L4
{"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48}
0.027261
16,777,216
360,448
46.545455
2
{"M": 256, "N": 64, "K": 512, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float16, device='cuda') B = torch.randn(512, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA L40S
{"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96}
0.057054
16,777,216
360,448
46.545455
2
{"M": 256, "N": 64, "K": 512, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float16, device='cuda') B = torch.randn(512, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX 3090
{"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6}
0.020788
16,777,216
360,448
46.545455
2
{"M": 256, "N": 64, "K": 512, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float16, device='cuda') B = torch.randn(512, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX 4090
{"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72}
0.044643
16,777,216
360,448
46.545455
2
{"M": 256, "N": 64, "K": 512, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float16, device='cuda') B = torch.randn(512, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA H100 SXM
{"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50}
0.040593
16,777,216
360,448
46.545455
2
{"M": 256, "N": 64, "K": 512, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float16, device='cuda') B = torch.randn(512, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA H100 PCIe
{"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50}
0.05702
16,777,216
360,448
46.545455
2
{"M": 256, "N": 64, "K": 512, "dtype": "float16"}
import torch def matmul_kernel(A, B): # Matrix multiplication: (256, 512) x (512, 64) -> (256, 64) C = torch.matmul(A, B) return C A = torch.randn(256, 512, dtype=torch.float16, device='cuda') B = torch.randn(512, 64, dtype=torch.float16, device='cuda') C = matmul_kernel(A, B) torch.cuda.synchronize()
matmul
NVIDIA RTX A6000
{"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6}
0.040461
16,777,216
360,448
46.545455
2
{"M": 256, "N": 64, "K": 512, "dtype": "float16"}