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: (2048, 256) x (256, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 256, 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} | 0.048893 | 268,435,456 | 2,228,224 | 120.470588 | 2 | {"M": 2048, "N": 256, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 256, 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} | 0.057102 | 268,435,456 | 2,228,224 | 120.470588 | 2 | {"M": 2048, "N": 256, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 256, 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} | 0.029388 | 268,435,456 | 2,228,224 | 120.470588 | 2 | {"M": 2048, "N": 256, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 256, 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} | 0.027801 | 268,435,456 | 2,228,224 | 120.470588 | 2 | {"M": 2048, "N": 256, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 256, 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} | 0.056607 | 268,435,456 | 2,228,224 | 120.470588 | 2 | {"M": 2048, "N": 256, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 256, 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} | 0.062363 | 268,435,456 | 2,228,224 | 120.470588 | 2 | {"M": 2048, "N": 256, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 256, 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} | 0.056943 | 268,435,456 | 2,228,224 | 120.470588 | 2 | {"M": 2048, "N": 256, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 256, 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} | 0.064837 | 268,435,456 | 2,228,224 | 120.470588 | 2 | {"M": 2048, "N": 256, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 256, 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} | 0.060039 | 268,435,456 | 2,228,224 | 120.470588 | 2 | {"M": 2048, "N": 256, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 256, 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} | 0.03618 | 268,435,456 | 2,228,224 | 120.470588 | 2 | {"M": 2048, "N": 256, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 256, 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} | 0.034709 | 268,435,456 | 2,228,224 | 120.470588 | 2 | {"M": 2048, "N": 256, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 256, 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} | 0.055964 | 268,435,456 | 2,228,224 | 120.470588 | 2 | {"M": 2048, "N": 256, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 256, 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} | 0.115105 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 256, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 256, 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} | 0.075325 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 256, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 256, 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} | 0.056779 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 256, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 256, 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} | 0.082869 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 256, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 256, 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} | 0.067917 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 256, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 256, 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} | 0.051045 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 256, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 256, 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} | 0.053893 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 256, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 256, 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} | 0.050675 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 256, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 256, 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} | 0.039182 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 256, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 256, 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} | 0.047347 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 256, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 256, 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} | 0.053698 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 256, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 256, 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} | 0.070579 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 256, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 256, 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} | 0.054441 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 256, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 256, 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} | 0.073724 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 256, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 256, 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} | 0.05143 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 256, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 256, 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} | 0.062131 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 256, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 256, 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} | 0.046238 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 256, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 256, 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} | 0.061968 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 256, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 256, 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} | 0.069641 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 256, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 256, 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} | 0.060527 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 256, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 256, 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} | 0.032688 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 256, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 256, 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} | 0.061331 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 256, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 256, 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} | 0.028379 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 256, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 256, 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} | 0.058625 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 256, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.28208 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.182367 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.086794 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.121989 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.12893 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.096763 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.072931 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.065847 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.042264 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.073004 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.060515 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.09307 | 1,073,741,824 | 11,534,336 | 93.090909 | 4 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.064518 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.056885 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.047329 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.046424 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.05302 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.047568 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.056613 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.055539 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.036149 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.049873 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.054216 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.038609 | 1,073,741,824 | 5,767,168 | 186.181818 | 2 | {"M": 2048, "N": 256, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA 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} | 0.332792 | 2,147,483,648 | 20,971,520 | 102.4 | 4 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA 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} | 0.193551 | 2,147,483,648 | 20,971,520 | 102.4 | 4 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA 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} | 0.128719 | 2,147,483,648 | 20,971,520 | 102.4 | 4 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A100 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} | 0.163868 | 2,147,483,648 | 20,971,520 | 102.4 | 4 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.179194 | 2,147,483,648 | 20,971,520 | 102.4 | 4 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.131144 | 2,147,483,648 | 20,971,520 | 102.4 | 4 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.064174 | 2,147,483,648 | 20,971,520 | 102.4 | 4 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.152177 | 2,147,483,648 | 20,971,520 | 102.4 | 4 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.088399 | 2,147,483,648 | 20,971,520 | 102.4 | 4 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.063932 | 2,147,483,648 | 20,971,520 | 102.4 | 4 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.111411 | 2,147,483,648 | 20,971,520 | 102.4 | 4 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.098837 | 2,147,483,648 | 20,971,520 | 102.4 | 4 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.081055 | 2,147,483,648 | 10,485,760 | 204.8 | 2 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.072682 | 2,147,483,648 | 10,485,760 | 204.8 | 2 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.10058 | 2,147,483,648 | 10,485,760 | 204.8 | 2 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.059397 | 2,147,483,648 | 10,485,760 | 204.8 | 2 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.049075 | 2,147,483,648 | 10,485,760 | 204.8 | 2 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.091213 | 2,147,483,648 | 10,485,760 | 204.8 | 2 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.029553 | 2,147,483,648 | 10,485,760 | 204.8 | 2 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.081224 | 2,147,483,648 | 10,485,760 | 204.8 | 2 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.076592 | 2,147,483,648 | 10,485,760 | 204.8 | 2 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.044267 | 2,147,483,648 | 10,485,760 | 204.8 | 2 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.068281 | 2,147,483,648 | 10,485,760 | 204.8 | 2 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 2048) x (2048, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.079081 | 2,147,483,648 | 10,485,760 | 204.8 | 2 | {"M": 2048, "N": 256, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.803101 | 4,294,967,296 | 39,845,888 | 107.789474 | 4 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.372992 | 4,294,967,296 | 39,845,888 | 107.789474 | 4 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.201735 | 4,294,967,296 | 39,845,888 | 107.789474 | 4 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.263764 | 4,294,967,296 | 39,845,888 | 107.789474 | 4 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.289966 | 4,294,967,296 | 39,845,888 | 107.789474 | 4 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.241031 | 4,294,967,296 | 39,845,888 | 107.789474 | 4 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.078199 | 4,294,967,296 | 39,845,888 | 107.789474 | 4 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.215683 | 4,294,967,296 | 39,845,888 | 107.789474 | 4 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.100743 | 4,294,967,296 | 39,845,888 | 107.789474 | 4 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.090106 | 4,294,967,296 | 39,845,888 | 107.789474 | 4 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.14341 | 4,294,967,296 | 39,845,888 | 107.789474 | 4 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.169669 | 4,294,967,296 | 39,845,888 | 107.789474 | 4 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.141463 | 4,294,967,296 | 19,922,944 | 215.578947 | 2 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.112422 | 4,294,967,296 | 19,922,944 | 215.578947 | 2 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.111645 | 4,294,967,296 | 19,922,944 | 215.578947 | 2 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 4096) x (4096, 256) -> (2048, 256)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 256, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchro... | matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.068114 | 4,294,967,296 | 19,922,944 | 215.578947 | 2 | {"M": 2048, "N": 256, "K": 4096, "dtype": "float16"} |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.