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, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 2.029946 | 34,359,738,368 | 100,663,296 | 341.333333 | 4 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 1.542597 | 34,359,738,368 | 100,663,296 | 341.333333 | 4 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.647861 | 34,359,738,368 | 100,663,296 | 341.333333 | 4 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 1.4607 | 34,359,738,368 | 100,663,296 | 341.333333 | 4 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.5256 | 34,359,738,368 | 100,663,296 | 341.333333 | 4 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.580085 | 34,359,738,368 | 100,663,296 | 341.333333 | 4 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.986578 | 34,359,738,368 | 100,663,296 | 341.333333 | 4 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 1.343538 | 34,359,738,368 | 100,663,296 | 341.333333 | 4 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.586273 | 34,359,738,368 | 50,331,648 | 682.666667 | 2 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.555623 | 34,359,738,368 | 50,331,648 | 682.666667 | 2 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.723787 | 34,359,738,368 | 50,331,648 | 682.666667 | 2 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.192929 | 34,359,738,368 | 50,331,648 | 682.666667 | 2 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.158214 | 34,359,738,368 | 50,331,648 | 682.666667 | 2 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.405895 | 34,359,738,368 | 50,331,648 | 682.666667 | 2 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.237516 | 34,359,738,368 | 50,331,648 | 682.666667 | 2 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.583396 | 34,359,738,368 | 50,331,648 | 682.666667 | 2 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.393156 | 34,359,738,368 | 50,331,648 | 682.666667 | 2 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.108465 | 34,359,738,368 | 50,331,648 | 682.666667 | 2 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.148581 | 34,359,738,368 | 50,331,648 | 682.666667 | 2 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (1024, 4096) x (4096, 4096) -> (1024, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.sync... | 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} | 0.717156 | 34,359,738,368 | 50,331,648 | 682.666667 | 2 | {"M": 1024, "N": 4096, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.029736 | 16,777,216 | 1,064,960 | 15.753846 | 4 | {"M": 2048, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.04687 | 16,777,216 | 1,064,960 | 15.753846 | 4 | {"M": 2048, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.035439 | 16,777,216 | 1,064,960 | 15.753846 | 4 | {"M": 2048, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.058892 | 16,777,216 | 1,064,960 | 15.753846 | 4 | {"M": 2048, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.032604 | 16,777,216 | 1,064,960 | 15.753846 | 4 | {"M": 2048, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.037449 | 16,777,216 | 1,064,960 | 15.753846 | 4 | {"M": 2048, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.064254 | 16,777,216 | 1,064,960 | 15.753846 | 4 | {"M": 2048, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.033166 | 16,777,216 | 1,064,960 | 15.753846 | 4 | {"M": 2048, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.041644 | 16,777,216 | 1,064,960 | 15.753846 | 4 | {"M": 2048, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.036286 | 16,777,216 | 1,064,960 | 15.753846 | 4 | {"M": 2048, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.058647 | 16,777,216 | 1,064,960 | 15.753846 | 4 | {"M": 2048, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.030226 | 16,777,216 | 1,064,960 | 15.753846 | 4 | {"M": 2048, "N": 64, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.022488 | 16,777,216 | 532,480 | 31.507692 | 2 | {"M": 2048, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.03171 | 16,777,216 | 532,480 | 31.507692 | 2 | {"M": 2048, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.040732 | 16,777,216 | 532,480 | 31.507692 | 2 | {"M": 2048, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.040443 | 16,777,216 | 532,480 | 31.507692 | 2 | {"M": 2048, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.030583 | 16,777,216 | 532,480 | 31.507692 | 2 | {"M": 2048, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.052718 | 16,777,216 | 532,480 | 31.507692 | 2 | {"M": 2048, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.041143 | 16,777,216 | 532,480 | 31.507692 | 2 | {"M": 2048, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.04997 | 16,777,216 | 532,480 | 31.507692 | 2 | {"M": 2048, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.058162 | 16,777,216 | 532,480 | 31.507692 | 2 | {"M": 2048, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.041926 | 16,777,216 | 532,480 | 31.507692 | 2 | {"M": 2048, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.038293 | 16,777,216 | 532,480 | 31.507692 | 2 | {"M": 2048, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.054748 | 16,777,216 | 532,480 | 31.507692 | 2 | {"M": 2048, "N": 64, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.047486 | 67,108,864 | 2,686,976 | 24.97561 | 4 | {"M": 2048, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.03314 | 67,108,864 | 2,686,976 | 24.97561 | 4 | {"M": 2048, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.068132 | 67,108,864 | 2,686,976 | 24.97561 | 4 | {"M": 2048, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.056996 | 67,108,864 | 2,686,976 | 24.97561 | 4 | {"M": 2048, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.037998 | 67,108,864 | 2,686,976 | 24.97561 | 4 | {"M": 2048, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.039193 | 67,108,864 | 2,686,976 | 24.97561 | 4 | {"M": 2048, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.059676 | 67,108,864 | 2,686,976 | 24.97561 | 4 | {"M": 2048, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.027269 | 67,108,864 | 2,686,976 | 24.97561 | 4 | {"M": 2048, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.039616 | 67,108,864 | 2,686,976 | 24.97561 | 4 | {"M": 2048, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.039753 | 67,108,864 | 2,686,976 | 24.97561 | 4 | {"M": 2048, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.026878 | 67,108,864 | 2,686,976 | 24.97561 | 4 | {"M": 2048, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.055098 | 67,108,864 | 2,686,976 | 24.97561 | 4 | {"M": 2048, "N": 64, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.054544 | 67,108,864 | 1,343,488 | 49.95122 | 2 | {"M": 2048, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.051991 | 67,108,864 | 1,343,488 | 49.95122 | 2 | {"M": 2048, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.053013 | 67,108,864 | 1,343,488 | 49.95122 | 2 | {"M": 2048, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.030404 | 67,108,864 | 1,343,488 | 49.95122 | 2 | {"M": 2048, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.053785 | 67,108,864 | 1,343,488 | 49.95122 | 2 | {"M": 2048, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.027307 | 67,108,864 | 1,343,488 | 49.95122 | 2 | {"M": 2048, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.030761 | 67,108,864 | 1,343,488 | 49.95122 | 2 | {"M": 2048, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 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} | 0.054304 | 67,108,864 | 1,343,488 | 49.95122 | 2 | {"M": 2048, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.072954 | 67,108,864 | 1,343,488 | 49.95122 | 2 | {"M": 2048, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.033121 | 67,108,864 | 1,343,488 | 49.95122 | 2 | {"M": 2048, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.026927 | 67,108,864 | 1,343,488 | 49.95122 | 2 | {"M": 2048, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.048658 | 67,108,864 | 1,343,488 | 49.95122 | 2 | {"M": 2048, "N": 64, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.075303 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 64, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.046034 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 64, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.060557 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 64, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.066664 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 64, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.046808 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 64, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.073187 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 64, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.042674 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 64, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.033812 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 64, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.061886 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 64, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.039794 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 64, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.059461 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 64, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.057172 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 64, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.044969 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 64, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.031199 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 64, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.047917 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 64, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.040167 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 64, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.045695 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 64, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.057072 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 64, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.026873 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 64, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.05092 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 64, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.029924 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 64, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.073314 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 64, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.052365 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 64, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 64, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.057629 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 64, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchroniz... | 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} | 0.093009 | 268,435,456 | 9,175,040 | 29.257143 | 4 | {"M": 2048, "N": 64, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchroniz... | 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} | 0.058009 | 268,435,456 | 9,175,040 | 29.257143 | 4 | {"M": 2048, "N": 64, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchroniz... | 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} | 0.078977 | 268,435,456 | 9,175,040 | 29.257143 | 4 | {"M": 2048, "N": 64, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchroniz... | 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} | 0.068878 | 268,435,456 | 9,175,040 | 29.257143 | 4 | {"M": 2048, "N": 64, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchroniz... | 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} | 0.07452 | 268,435,456 | 9,175,040 | 29.257143 | 4 | {"M": 2048, "N": 64, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchroniz... | 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} | 0.045941 | 268,435,456 | 9,175,040 | 29.257143 | 4 | {"M": 2048, "N": 64, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchroniz... | 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} | 0.062564 | 268,435,456 | 9,175,040 | 29.257143 | 4 | {"M": 2048, "N": 64, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 64) -> (2048, 64)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 64, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchroniz... | 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} | 0.065697 | 268,435,456 | 9,175,040 | 29.257143 | 4 | {"M": 2048, "N": 64, "K": 1024, "dtype": "float32"} |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.