code large_stringlengths 193 899 | workload_type large_stringclasses 15
values | gpu_name large_stringclasses 12
values | gpu_features large_stringclasses 12
values | runtime_ms float64 0.01 4.35k | flops float64 30 21,045B | memory_bytes int64 160 13.2B | arithmetic_intensity float64 0 6.37k | dtype_bytes int64 2 4 | workload_params large_stringlengths 11 79 |
|---|---|---|---|---|---|---|---|---|---|
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.092029 | 1,073,741,824 | 8,388,608 | 128 | 4 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.130168 | 1,073,741,824 | 8,388,608 | 128 | 4 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.054183 | 1,073,741,824 | 8,388,608 | 128 | 4 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.087859 | 1,073,741,824 | 8,388,608 | 128 | 4 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.048122 | 1,073,741,824 | 8,388,608 | 128 | 4 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.040213 | 1,073,741,824 | 8,388,608 | 128 | 4 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.066151 | 1,073,741,824 | 8,388,608 | 128 | 4 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.086889 | 1,073,741,824 | 8,388,608 | 128 | 4 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.070338 | 1,073,741,824 | 4,194,304 | 256 | 2 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.049985 | 1,073,741,824 | 4,194,304 | 256 | 2 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.05413 | 1,073,741,824 | 4,194,304 | 256 | 2 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.055557 | 1,073,741,824 | 4,194,304 | 256 | 2 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.048735 | 1,073,741,824 | 4,194,304 | 256 | 2 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.069717 | 1,073,741,824 | 4,194,304 | 256 | 2 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.06891 | 1,073,741,824 | 4,194,304 | 256 | 2 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.066811 | 1,073,741,824 | 4,194,304 | 256 | 2 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.070381 | 1,073,741,824 | 4,194,304 | 256 | 2 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.025074 | 1,073,741,824 | 4,194,304 | 256 | 2 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.064041 | 1,073,741,824 | 4,194,304 | 256 | 2 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 1024) x (1024, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 1024, 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} | 0.052766 | 1,073,741,824 | 4,194,304 | 256 | 2 | {"M": 512, "N": 1024, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.420556 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.267138 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.124596 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.164418 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.184391 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.130527 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.066179 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.153792 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.081317 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.092772 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.09344 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float32, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.139751 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.086592 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.084282 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.081021 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.068677 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.042959 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.07737 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.061582 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.080716 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.081972 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.050214 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.045789 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 2048) x (2048, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 2048, dtype=torch.float16, device='cuda')
B = torch.randn(2048, 1024, 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} | 0.061918 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 512, "N": 1024, "K": 2048, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.876478 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.390784 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.198062 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.351036 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.369453 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.167463 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.115346 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.181624 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.100204 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.122612 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.159959 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float32, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.203404 | 4,294,967,296 | 27,262,976 | 157.538462 | 4 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.148803 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.130832 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.163057 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.058164 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.070916 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.091488 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.092487 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.138265 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.060358 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.056537 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.047354 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 4096) x (4096, 1024) -> (512, 1024)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 4096, dtype=torch.float16, device='cuda')
B = torch.randn(4096, 1024, 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} | 0.143799 | 4,294,967,296 | 13,631,488 | 315.076923 | 2 | {"M": 512, "N": 1024, "K": 4096, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 2048, 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} | 0.079223 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 512, "N": 2048, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 2048, 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} | 0.073889 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 512, "N": 2048, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 2048, 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} | 0.069828 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 512, "N": 2048, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 2048, 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} | 0.054108 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 512, "N": 2048, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 2048, 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} | 0.056507 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 512, "N": 2048, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 2048, 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} | 0.03238 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 512, "N": 2048, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 2048, 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} | 0.040922 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 512, "N": 2048, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 2048, 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} | 0.037879 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 512, "N": 2048, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 2048, 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} | 0.069444 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 512, "N": 2048, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 2048, 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} | 0.075091 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 512, "N": 2048, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 2048, 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} | 0.031525 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 512, "N": 2048, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 2048, 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} | 0.051191 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 512, "N": 2048, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 2048, 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} | 0.07729 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 512, "N": 2048, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 2048, 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} | 0.033079 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 512, "N": 2048, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 2048, 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} | 0.059354 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 512, "N": 2048, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 2048, 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} | 0.039986 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 512, "N": 2048, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 2048, 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} | 0.048763 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 512, "N": 2048, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 2048, 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} | 0.072603 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 512, "N": 2048, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 2048, 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} | 0.024368 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 512, "N": 2048, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 2048, 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} | 0.022617 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 512, "N": 2048, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 2048, 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} | 0.046738 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 512, "N": 2048, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 2048, 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} | 0.051618 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 512, "N": 2048, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 2048, 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} | 0.049636 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 512, "N": 2048, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 64) x (64, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 2048, 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} | 0.035352 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 512, "N": 2048, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 2048, 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} | 0.108951 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 512, "N": 2048, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 2048, 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} | 0.081941 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 512, "N": 2048, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 2048, 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} | 0.064746 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 512, "N": 2048, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 2048, 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} | 0.091041 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 512, "N": 2048, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 2048, 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} | 0.109387 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 512, "N": 2048, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 2048, 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} | 0.057541 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 512, "N": 2048, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 2048, 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} | 0.047076 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 512, "N": 2048, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (512, 256) x (256, 2048) -> (512, 2048)
C = torch.matmul(A, B)
return C
A = torch.randn(512, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 2048, 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} | 0.086105 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 512, "N": 2048, "K": 256, "dtype": "float32"} |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.