code large_stringlengths 193 899 | workload_type large_stringclasses 15
values | gpu_name large_stringclasses 12
values | gpu_features large_stringclasses 12
values | runtime_ms float64 0.01 4.35k | flops float64 30 21,045B | memory_bytes int64 160 13.2B | arithmetic_intensity float64 0 6.37k | dtype_bytes int64 2 4 | workload_params large_stringlengths 11 79 |
|---|---|---|---|---|---|---|---|---|---|
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 2048, 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} | 0.222883 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 2048, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 2048, 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} | 0.122655 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 2048, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 2048, 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} | 0.250492 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 2048, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 2048, 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} | 0.100816 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 2048, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 2048, 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} | 0.073699 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 2048, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 2048, 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} | 0.161061 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 2048, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 2048, 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} | 0.093154 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 2048, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 2048, 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} | 0.158419 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 2048, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 2048, 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} | 0.129975 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 2048, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 2048, 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} | 0.024809 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 2048, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 2048, 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} | 0.066054 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 2048, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 2048, 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} | 0.176224 | 8,589,934,592 | 23,068,672 | 372.363636 | 2 | {"M": 512, "N": 2048, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 4096, 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} | 0.118442 | 268,435,456 | 9,568,256 | 28.054795 | 4 | {"M": 512, "N": 4096, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 4096, 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} | 0.068452 | 268,435,456 | 9,568,256 | 28.054795 | 4 | {"M": 512, "N": 4096, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 4096, 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} | 0.073198 | 268,435,456 | 9,568,256 | 28.054795 | 4 | {"M": 512, "N": 4096, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 4096, 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} | 0.063542 | 268,435,456 | 9,568,256 | 28.054795 | 4 | {"M": 512, "N": 4096, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 4096, 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} | 0.066054 | 268,435,456 | 9,568,256 | 28.054795 | 4 | {"M": 512, "N": 4096, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 4096, 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} | 0.066591 | 268,435,456 | 9,568,256 | 28.054795 | 4 | {"M": 512, "N": 4096, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 4096, 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} | 0.041869 | 268,435,456 | 9,568,256 | 28.054795 | 4 | {"M": 512, "N": 4096, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 4096, 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} | 0.062925 | 268,435,456 | 9,568,256 | 28.054795 | 4 | {"M": 512, "N": 4096, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 4096, 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} | 0.062588 | 268,435,456 | 9,568,256 | 28.054795 | 4 | {"M": 512, "N": 4096, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 4096, 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} | 0.059116 | 268,435,456 | 9,568,256 | 28.054795 | 4 | {"M": 512, "N": 4096, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 4096, 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} | 0.055203 | 268,435,456 | 9,568,256 | 28.054795 | 4 | {"M": 512, "N": 4096, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 4096, 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} | 0.061294 | 268,435,456 | 9,568,256 | 28.054795 | 4 | {"M": 512, "N": 4096, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 4096, 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} | 0.083208 | 268,435,456 | 4,784,128 | 56.109589 | 2 | {"M": 512, "N": 4096, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 4096, 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} | 0.022337 | 268,435,456 | 4,784,128 | 56.109589 | 2 | {"M": 512, "N": 4096, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 4096, 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} | 0.059885 | 268,435,456 | 4,784,128 | 56.109589 | 2 | {"M": 512, "N": 4096, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 4096, 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} | 0.029511 | 268,435,456 | 4,784,128 | 56.109589 | 2 | {"M": 512, "N": 4096, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 4096, 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} | 0.065457 | 268,435,456 | 4,784,128 | 56.109589 | 2 | {"M": 512, "N": 4096, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 4096, 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} | 0.056798 | 268,435,456 | 4,784,128 | 56.109589 | 2 | {"M": 512, "N": 4096, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 4096, 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} | 0.033261 | 268,435,456 | 4,784,128 | 56.109589 | 2 | {"M": 512, "N": 4096, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 4096, 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} | 0.059331 | 268,435,456 | 4,784,128 | 56.109589 | 2 | {"M": 512, "N": 4096, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 4096, 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} | 0.056753 | 268,435,456 | 4,784,128 | 56.109589 | 2 | {"M": 512, "N": 4096, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 4096, 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} | 0.059161 | 268,435,456 | 4,784,128 | 56.109589 | 2 | {"M": 512, "N": 4096, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 4096, 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} | 0.049994 | 268,435,456 | 4,784,128 | 56.109589 | 2 | {"M": 512, "N": 4096, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 4096, 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} | 0.035747 | 268,435,456 | 4,784,128 | 56.109589 | 2 | {"M": 512, "N": 4096, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 4096, 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} | 0.22069 | 1,073,741,824 | 13,107,200 | 81.92 | 4 | {"M": 512, "N": 4096, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 4096, 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} | 0.131009 | 1,073,741,824 | 13,107,200 | 81.92 | 4 | {"M": 512, "N": 4096, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 4096, 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} | 0.094744 | 1,073,741,824 | 13,107,200 | 81.92 | 4 | {"M": 512, "N": 4096, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 4096, 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} | 0.124524 | 1,073,741,824 | 13,107,200 | 81.92 | 4 | {"M": 512, "N": 4096, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 4096, 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} | 0.12689 | 1,073,741,824 | 13,107,200 | 81.92 | 4 | {"M": 512, "N": 4096, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 4096, 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} | 0.060731 | 1,073,741,824 | 13,107,200 | 81.92 | 4 | {"M": 512, "N": 4096, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 4096, 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} | 0.062649 | 1,073,741,824 | 13,107,200 | 81.92 | 4 | {"M": 512, "N": 4096, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 4096, 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} | 0.066789 | 1,073,741,824 | 13,107,200 | 81.92 | 4 | {"M": 512, "N": 4096, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 4096, 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} | 0.069802 | 1,073,741,824 | 13,107,200 | 81.92 | 4 | {"M": 512, "N": 4096, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 4096, 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} | 0.073951 | 1,073,741,824 | 13,107,200 | 81.92 | 4 | {"M": 512, "N": 4096, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 4096, 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} | 0.090046 | 1,073,741,824 | 13,107,200 | 81.92 | 4 | {"M": 512, "N": 4096, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 4096, 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} | 0.072575 | 1,073,741,824 | 13,107,200 | 81.92 | 4 | {"M": 512, "N": 4096, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 4096, 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} | 0.056937 | 1,073,741,824 | 6,553,600 | 163.84 | 2 | {"M": 512, "N": 4096, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 4096, 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} | 0.057843 | 1,073,741,824 | 6,553,600 | 163.84 | 2 | {"M": 512, "N": 4096, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 4096, 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} | 0.080403 | 1,073,741,824 | 6,553,600 | 163.84 | 2 | {"M": 512, "N": 4096, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 4096, 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} | 0.034923 | 1,073,741,824 | 6,553,600 | 163.84 | 2 | {"M": 512, "N": 4096, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 4096, 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} | 0.045448 | 1,073,741,824 | 6,553,600 | 163.84 | 2 | {"M": 512, "N": 4096, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 4096, 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} | 0.043464 | 1,073,741,824 | 6,553,600 | 163.84 | 2 | {"M": 512, "N": 4096, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 4096, 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} | 0.039342 | 1,073,741,824 | 6,553,600 | 163.84 | 2 | {"M": 512, "N": 4096, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 4096, 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} | 0.044532 | 1,073,741,824 | 6,553,600 | 163.84 | 2 | {"M": 512, "N": 4096, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 4096, 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} | 0.066589 | 1,073,741,824 | 6,553,600 | 163.84 | 2 | {"M": 512, "N": 4096, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 4096, 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} | 0.034721 | 1,073,741,824 | 6,553,600 | 163.84 | 2 | {"M": 512, "N": 4096, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 4096, 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} | 0.072207 | 1,073,741,824 | 6,553,600 | 163.84 | 2 | {"M": 512, "N": 4096, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 4096, 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} | 0.059212 | 1,073,741,824 | 6,553,600 | 163.84 | 2 | {"M": 512, "N": 4096, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 4096, 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} | 0.44414 | 2,147,483,648 | 17,825,792 | 120.470588 | 4 | {"M": 512, "N": 4096, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 4096, 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} | 0.240996 | 2,147,483,648 | 17,825,792 | 120.470588 | 4 | {"M": 512, "N": 4096, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 4096, 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} | 0.136054 | 2,147,483,648 | 17,825,792 | 120.470588 | 4 | {"M": 512, "N": 4096, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 4096, 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} | 0.176706 | 2,147,483,648 | 17,825,792 | 120.470588 | 4 | {"M": 512, "N": 4096, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 4096, 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} | 0.173811 | 2,147,483,648 | 17,825,792 | 120.470588 | 4 | {"M": 512, "N": 4096, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 4096, 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} | 0.162955 | 2,147,483,648 | 17,825,792 | 120.470588 | 4 | {"M": 512, "N": 4096, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 4096, 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} | 0.086246 | 2,147,483,648 | 17,825,792 | 120.470588 | 4 | {"M": 512, "N": 4096, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 4096, 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} | 0.134596 | 2,147,483,648 | 17,825,792 | 120.470588 | 4 | {"M": 512, "N": 4096, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 4096, 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} | 0.090183 | 2,147,483,648 | 17,825,792 | 120.470588 | 4 | {"M": 512, "N": 4096, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 4096, 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} | 0.078521 | 2,147,483,648 | 17,825,792 | 120.470588 | 4 | {"M": 512, "N": 4096, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 4096, 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} | 0.092446 | 2,147,483,648 | 17,825,792 | 120.470588 | 4 | {"M": 512, "N": 4096, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 4096, 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} | 0.15481 | 2,147,483,648 | 17,825,792 | 120.470588 | 4 | {"M": 512, "N": 4096, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 4096, 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} | 0.065284 | 2,147,483,648 | 8,912,896 | 240.941176 | 2 | {"M": 512, "N": 4096, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 4096, 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} | 0.074524 | 2,147,483,648 | 8,912,896 | 240.941176 | 2 | {"M": 512, "N": 4096, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 4096, 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} | 0.089947 | 2,147,483,648 | 8,912,896 | 240.941176 | 2 | {"M": 512, "N": 4096, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 4096, 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} | 0.055994 | 2,147,483,648 | 8,912,896 | 240.941176 | 2 | {"M": 512, "N": 4096, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 4096, 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} | 0.038934 | 2,147,483,648 | 8,912,896 | 240.941176 | 2 | {"M": 512, "N": 4096, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 4096, 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} | 0.043093 | 2,147,483,648 | 8,912,896 | 240.941176 | 2 | {"M": 512, "N": 4096, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 4096, 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} | 0.040612 | 2,147,483,648 | 8,912,896 | 240.941176 | 2 | {"M": 512, "N": 4096, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 4096, 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} | 0.095481 | 2,147,483,648 | 8,912,896 | 240.941176 | 2 | {"M": 512, "N": 4096, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 4096, 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} | 0.049873 | 2,147,483,648 | 8,912,896 | 240.941176 | 2 | {"M": 512, "N": 4096, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 4096, 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} | 0.030516 | 2,147,483,648 | 8,912,896 | 240.941176 | 2 | {"M": 512, "N": 4096, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 4096, 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} | 0.065314 | 2,147,483,648 | 8,912,896 | 240.941176 | 2 | {"M": 512, "N": 4096, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 512) x (512, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 4096, 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} | 0.070199 | 2,147,483,648 | 8,912,896 | 240.941176 | 2 | {"M": 512, "N": 4096, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.973045 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.340481 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.299293 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.469063 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.369382 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.193254 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.094616 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.219347 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.090834 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.093724 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.155969 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.185416 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA 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} | 0.130421 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA 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} | 0.082234 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA 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} | 0.122211 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 4096) -> (512, 4096)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 4096, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA 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} | 0.051389 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 4096, "K": 1024, "dtype": "float16"} |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.