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: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 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} | 0.052716 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 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} | 0.085146 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 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} | 0.069797 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 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} | 0.13808 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 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} | 0.090652 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 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} | 0.043627 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 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} | 0.040147 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 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} | 0.143248 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 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} | 1.905242 | 8,589,934,592 | 46,137,344 | 186.181818 | 4 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 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} | 0.652554 | 8,589,934,592 | 46,137,344 | 186.181818 | 4 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 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} | 0.502355 | 8,589,934,592 | 46,137,344 | 186.181818 | 4 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 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} | 0.624389 | 8,589,934,592 | 46,137,344 | 186.181818 | 4 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 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} | 0.659665 | 8,589,934,592 | 46,137,344 | 186.181818 | 4 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 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} | 0.399578 | 8,589,934,592 | 46,137,344 | 186.181818 | 4 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 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} | 0.18085 | 8,589,934,592 | 46,137,344 | 186.181818 | 4 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 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} | 0.382942 | 8,589,934,592 | 46,137,344 | 186.181818 | 4 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 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} | 0.139217 | 8,589,934,592 | 46,137,344 | 186.181818 | 4 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 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} | 0.239284 | 8,589,934,592 | 46,137,344 | 186.181818 | 4 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 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} | 0.321486 | 8,589,934,592 | 46,137,344 | 186.181818 | 4 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 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} | 0.259279 | 8,589,934,592 | 46,137,344 | 186.181818 | 4 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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 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} | 0.201321 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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 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} | 0.170884 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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 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} | 0.191415 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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 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} | 0.100106 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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 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} | 0.080258 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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 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} | 0.085826 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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 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} | 0.106588 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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 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} | 0.200139 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.091583 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.064439 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.058407 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.184581 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 4096, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 3.53912 | 17,179,869,184 | 83,886,080 | 204.8 | 4 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 1.354177 | 17,179,869,184 | 83,886,080 | 204.8 | 4 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.699907 | 17,179,869,184 | 83,886,080 | 204.8 | 4 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 1.124998 | 17,179,869,184 | 83,886,080 | 204.8 | 4 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 1.093459 | 17,179,869,184 | 83,886,080 | 204.8 | 4 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.776179 | 17,179,869,184 | 83,886,080 | 204.8 | 4 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.210003 | 17,179,869,184 | 83,886,080 | 204.8 | 4 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.633245 | 17,179,869,184 | 83,886,080 | 204.8 | 4 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.364965 | 17,179,869,184 | 83,886,080 | 204.8 | 4 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.396557 | 17,179,869,184 | 83,886,080 | 204.8 | 4 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.457693 | 17,179,869,184 | 83,886,080 | 204.8 | 4 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.572411 | 17,179,869,184 | 83,886,080 | 204.8 | 4 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.286754 | 17,179,869,184 | 41,943,040 | 409.6 | 2 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.23258 | 17,179,869,184 | 41,943,040 | 409.6 | 2 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.410476 | 17,179,869,184 | 41,943,040 | 409.6 | 2 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.129385 | 17,179,869,184 | 41,943,040 | 409.6 | 2 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.103061 | 17,179,869,184 | 41,943,040 | 409.6 | 2 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.196189 | 17,179,869,184 | 41,943,040 | 409.6 | 2 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.189781 | 17,179,869,184 | 41,943,040 | 409.6 | 2 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.321736 | 17,179,869,184 | 41,943,040 | 409.6 | 2 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.186798 | 17,179,869,184 | 41,943,040 | 409.6 | 2 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.073914 | 17,179,869,184 | 41,943,040 | 409.6 | 2 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.0801 | 17,179,869,184 | 41,943,040 | 409.6 | 2 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 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} | 0.408331 | 17,179,869,184 | 41,943,040 | 409.6 | 2 | {"M": 512, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.032961 | 8,388,608 | 540,672 | 15.515152 | 4 | {"M": 1024, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.027237 | 8,388,608 | 540,672 | 15.515152 | 4 | {"M": 1024, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.059972 | 8,388,608 | 540,672 | 15.515152 | 4 | {"M": 1024, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.04852 | 8,388,608 | 540,672 | 15.515152 | 4 | {"M": 1024, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.052932 | 8,388,608 | 540,672 | 15.515152 | 4 | {"M": 1024, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.062706 | 8,388,608 | 540,672 | 15.515152 | 4 | {"M": 1024, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.041378 | 8,388,608 | 540,672 | 15.515152 | 4 | {"M": 1024, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.042535 | 8,388,608 | 540,672 | 15.515152 | 4 | {"M": 1024, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.031157 | 8,388,608 | 540,672 | 15.515152 | 4 | {"M": 1024, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.049199 | 8,388,608 | 540,672 | 15.515152 | 4 | {"M": 1024, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.061575 | 8,388,608 | 540,672 | 15.515152 | 4 | {"M": 1024, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.027223 | 8,388,608 | 540,672 | 15.515152 | 4 | {"M": 1024, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.037837 | 8,388,608 | 270,336 | 31.030303 | 2 | {"M": 1024, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.047669 | 8,388,608 | 270,336 | 31.030303 | 2 | {"M": 1024, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.045876 | 8,388,608 | 270,336 | 31.030303 | 2 | {"M": 1024, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.034307 | 8,388,608 | 270,336 | 31.030303 | 2 | {"M": 1024, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.028736 | 8,388,608 | 270,336 | 31.030303 | 2 | {"M": 1024, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.031099 | 8,388,608 | 270,336 | 31.030303 | 2 | {"M": 1024, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.056461 | 8,388,608 | 270,336 | 31.030303 | 2 | {"M": 1024, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.04622 | 8,388,608 | 270,336 | 31.030303 | 2 | {"M": 1024, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.029934 | 8,388,608 | 270,336 | 31.030303 | 2 | {"M": 1024, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.030615 | 8,388,608 | 270,336 | 31.030303 | 2 | {"M": 1024, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.030698 | 8,388,608 | 270,336 | 31.030303 | 2 | {"M": 1024, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.045467 | 8,388,608 | 270,336 | 31.030303 | 2 | {"M": 1024, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.034977 | 33,554,432 | 1,376,256 | 24.380952 | 4 | {"M": 1024, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.02954 | 33,554,432 | 1,376,256 | 24.380952 | 4 | {"M": 1024, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.05312 | 33,554,432 | 1,376,256 | 24.380952 | 4 | {"M": 1024, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.025692 | 33,554,432 | 1,376,256 | 24.380952 | 4 | {"M": 1024, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.060689 | 33,554,432 | 1,376,256 | 24.380952 | 4 | {"M": 1024, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.047354 | 33,554,432 | 1,376,256 | 24.380952 | 4 | {"M": 1024, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.074376 | 33,554,432 | 1,376,256 | 24.380952 | 4 | {"M": 1024, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.050701 | 33,554,432 | 1,376,256 | 24.380952 | 4 | {"M": 1024, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.052646 | 33,554,432 | 1,376,256 | 24.380952 | 4 | {"M": 1024, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.048612 | 33,554,432 | 1,376,256 | 24.380952 | 4 | {"M": 1024, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.037462 | 33,554,432 | 1,376,256 | 24.380952 | 4 | {"M": 1024, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.061084 | 33,554,432 | 1,376,256 | 24.380952 | 4 | {"M": 1024, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.067589 | 33,554,432 | 688,128 | 48.761905 | 2 | {"M": 1024, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.032814 | 33,554,432 | 688,128 | 48.761905 | 2 | {"M": 1024, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.031885 | 33,554,432 | 688,128 | 48.761905 | 2 | {"M": 1024, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.04569 | 33,554,432 | 688,128 | 48.761905 | 2 | {"M": 1024, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.03615 | 33,554,432 | 688,128 | 48.761905 | 2 | {"M": 1024, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.028722 | 33,554,432 | 688,128 | 48.761905 | 2 | {"M": 1024, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.038447 | 33,554,432 | 688,128 | 48.761905 | 2 | {"M": 1024, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 64) -> (1024, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 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} | 0.070065 | 33,554,432 | 688,128 | 48.761905 | 2 | {"M": 1024, "N": 64, "K": 256, "dtype": "float16"} |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.