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: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, 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} | 0.10481 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, 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} | 0.10966 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, 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} | 0.036216 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, 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} | 0.085434 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, 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} | 0.083372 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, 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} | 0.062304 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, 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} | 0.061756 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, 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} | 0.058087 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, 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} | 0.0653 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, 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} | 0.046282 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, 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} | 0.072599 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, 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} | 0.046756 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, 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} | 0.069133 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, 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} | 0.055628 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, 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} | 0.042171 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, 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} | 0.072067 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, 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} | 0.047548 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, 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} | 0.037578 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, 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} | 0.046892 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 2048) x (2048, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, 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} | 0.072452 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 1024, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, 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} | 0.346852 | 2,147,483,648 | 22,020,096 | 97.52381 | 4 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, 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} | 0.241893 | 2,147,483,648 | 22,020,096 | 97.52381 | 4 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, 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} | 0.112758 | 2,147,483,648 | 22,020,096 | 97.52381 | 4 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, 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} | 0.196265 | 2,147,483,648 | 22,020,096 | 97.52381 | 4 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, 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} | 0.210124 | 2,147,483,648 | 22,020,096 | 97.52381 | 4 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, 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} | 0.131256 | 2,147,483,648 | 22,020,096 | 97.52381 | 4 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, 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} | 0.076364 | 2,147,483,648 | 22,020,096 | 97.52381 | 4 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, 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} | 0.11754 | 2,147,483,648 | 22,020,096 | 97.52381 | 4 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, 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} | 0.091774 | 2,147,483,648 | 22,020,096 | 97.52381 | 4 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, 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} | 0.087753 | 2,147,483,648 | 22,020,096 | 97.52381 | 4 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, 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} | 0.111897 | 2,147,483,648 | 22,020,096 | 97.52381 | 4 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, 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} | 0.143639 | 2,147,483,648 | 22,020,096 | 97.52381 | 4 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, 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} | 0.100047 | 2,147,483,648 | 11,010,048 | 195.047619 | 2 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, 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} | 0.073018 | 2,147,483,648 | 11,010,048 | 195.047619 | 2 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, 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} | 0.065293 | 2,147,483,648 | 11,010,048 | 195.047619 | 2 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, 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} | 0.069284 | 2,147,483,648 | 11,010,048 | 195.047619 | 2 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, 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} | 0.058436 | 2,147,483,648 | 11,010,048 | 195.047619 | 2 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, 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} | 0.077214 | 2,147,483,648 | 11,010,048 | 195.047619 | 2 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, 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} | 0.050985 | 2,147,483,648 | 11,010,048 | 195.047619 | 2 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, 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} | 0.072471 | 2,147,483,648 | 11,010,048 | 195.047619 | 2 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, 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} | 0.075416 | 2,147,483,648 | 11,010,048 | 195.047619 | 2 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, 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} | 0.036019 | 2,147,483,648 | 11,010,048 | 195.047619 | 2 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, 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} | 0.044275 | 2,147,483,648 | 11,010,048 | 195.047619 | 2 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 256) -> (1024, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, 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} | 0.081749 | 2,147,483,648 | 11,010,048 | 195.047619 | 2 | {"M": 1024, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, 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} | 0.054695 | 67,108,864 | 2,490,368 | 26.947368 | 4 | {"M": 1024, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, 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} | 0.046008 | 67,108,864 | 2,490,368 | 26.947368 | 4 | {"M": 1024, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, 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} | 0.064882 | 67,108,864 | 2,490,368 | 26.947368 | 4 | {"M": 1024, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, 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} | 0.032891 | 67,108,864 | 2,490,368 | 26.947368 | 4 | {"M": 1024, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, 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} | 0.04209 | 67,108,864 | 2,490,368 | 26.947368 | 4 | {"M": 1024, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, 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} | 0.026772 | 67,108,864 | 2,490,368 | 26.947368 | 4 | {"M": 1024, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, 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} | 0.028141 | 67,108,864 | 2,490,368 | 26.947368 | 4 | {"M": 1024, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, 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} | 0.03503 | 67,108,864 | 2,490,368 | 26.947368 | 4 | {"M": 1024, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, 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} | 0.05587 | 67,108,864 | 2,490,368 | 26.947368 | 4 | {"M": 1024, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, 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} | 0.04735 | 67,108,864 | 2,490,368 | 26.947368 | 4 | {"M": 1024, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, 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} | 0.027521 | 67,108,864 | 2,490,368 | 26.947368 | 4 | {"M": 1024, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, 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} | 0.050972 | 67,108,864 | 2,490,368 | 26.947368 | 4 | {"M": 1024, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, 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} | 0.022751 | 67,108,864 | 1,245,184 | 53.894737 | 2 | {"M": 1024, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, 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} | 0.053664 | 67,108,864 | 1,245,184 | 53.894737 | 2 | {"M": 1024, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, 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} | 0.044478 | 67,108,864 | 1,245,184 | 53.894737 | 2 | {"M": 1024, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, 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} | 0.034381 | 67,108,864 | 1,245,184 | 53.894737 | 2 | {"M": 1024, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, 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} | 0.055517 | 67,108,864 | 1,245,184 | 53.894737 | 2 | {"M": 1024, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, 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} | 0.073056 | 67,108,864 | 1,245,184 | 53.894737 | 2 | {"M": 1024, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, 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} | 0.049278 | 67,108,864 | 1,245,184 | 53.894737 | 2 | {"M": 1024, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, 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} | 0.054172 | 67,108,864 | 1,245,184 | 53.894737 | 2 | {"M": 1024, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, 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} | 0.050679 | 67,108,864 | 1,245,184 | 53.894737 | 2 | {"M": 1024, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, 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} | 0.030466 | 67,108,864 | 1,245,184 | 53.894737 | 2 | {"M": 1024, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, 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} | 0.049678 | 67,108,864 | 1,245,184 | 53.894737 | 2 | {"M": 1024, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 64) x (64, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, 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} | 0.04928 | 67,108,864 | 1,245,184 | 53.894737 | 2 | {"M": 1024, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, 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} | 0.109476 | 268,435,456 | 3,670,016 | 73.142857 | 4 | {"M": 1024, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, 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} | 0.055321 | 268,435,456 | 3,670,016 | 73.142857 | 4 | {"M": 1024, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, 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} | 0.04598 | 268,435,456 | 3,670,016 | 73.142857 | 4 | {"M": 1024, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, 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} | 0.057258 | 268,435,456 | 3,670,016 | 73.142857 | 4 | {"M": 1024, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, 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} | 0.049236 | 268,435,456 | 3,670,016 | 73.142857 | 4 | {"M": 1024, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, 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} | 0.054428 | 268,435,456 | 3,670,016 | 73.142857 | 4 | {"M": 1024, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, 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} | 0.060637 | 268,435,456 | 3,670,016 | 73.142857 | 4 | {"M": 1024, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, 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} | 0.059548 | 268,435,456 | 3,670,016 | 73.142857 | 4 | {"M": 1024, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, 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} | 0.03382 | 268,435,456 | 3,670,016 | 73.142857 | 4 | {"M": 1024, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, 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} | 0.044943 | 268,435,456 | 3,670,016 | 73.142857 | 4 | {"M": 1024, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, 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} | 0.047837 | 268,435,456 | 3,670,016 | 73.142857 | 4 | {"M": 1024, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, 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} | 0.063747 | 268,435,456 | 3,670,016 | 73.142857 | 4 | {"M": 1024, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, 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} | 0.040999 | 268,435,456 | 1,835,008 | 146.285714 | 2 | {"M": 1024, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, 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} | 0.023708 | 268,435,456 | 1,835,008 | 146.285714 | 2 | {"M": 1024, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, 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} | 0.056862 | 268,435,456 | 1,835,008 | 146.285714 | 2 | {"M": 1024, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, 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} | 0.024044 | 268,435,456 | 1,835,008 | 146.285714 | 2 | {"M": 1024, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, 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} | 0.05791 | 268,435,456 | 1,835,008 | 146.285714 | 2 | {"M": 1024, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, 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} | 0.023728 | 268,435,456 | 1,835,008 | 146.285714 | 2 | {"M": 1024, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, 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} | 0.051178 | 268,435,456 | 1,835,008 | 146.285714 | 2 | {"M": 1024, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, 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} | 0.056407 | 268,435,456 | 1,835,008 | 146.285714 | 2 | {"M": 1024, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, 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} | 0.060656 | 268,435,456 | 1,835,008 | 146.285714 | 2 | {"M": 1024, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, 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} | 0.043318 | 268,435,456 | 1,835,008 | 146.285714 | 2 | {"M": 1024, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, 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} | 0.028147 | 268,435,456 | 1,835,008 | 146.285714 | 2 | {"M": 1024, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 256) x (256, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, 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} | 0.049559 | 268,435,456 | 1,835,008 | 146.285714 | 2 | {"M": 1024, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 512) x (512, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, 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} | 0.132882 | 536,870,912 | 5,242,880 | 102.4 | 4 | {"M": 1024, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 512) x (512, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, 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} | 0.070373 | 536,870,912 | 5,242,880 | 102.4 | 4 | {"M": 1024, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 512) x (512, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, 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} | 0.044569 | 536,870,912 | 5,242,880 | 102.4 | 4 | {"M": 1024, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 512) x (512, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, 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} | 0.079937 | 536,870,912 | 5,242,880 | 102.4 | 4 | {"M": 1024, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 512) x (512, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, 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} | 0.091045 | 536,870,912 | 5,242,880 | 102.4 | 4 | {"M": 1024, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 512) x (512, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, 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} | 0.067627 | 536,870,912 | 5,242,880 | 102.4 | 4 | {"M": 1024, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 512) x (512, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, 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} | 0.058973 | 536,870,912 | 5,242,880 | 102.4 | 4 | {"M": 1024, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 512) x (512, 512) -> (1024, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, 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} | 0.04312 | 536,870,912 | 5,242,880 | 102.4 | 4 | {"M": 1024, "N": 512, "K": 512, "dtype": "float32"} |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.