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, 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 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} | 0.043471 | 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 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} | 0.10081 | 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 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} | 0.096207 | 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 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} | 0.100629 | 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 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} | 0.076581 | 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 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} | 0.055025 | 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 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} | 0.039376 | 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 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} | 0.122186 | 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, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.03559 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.048371 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.056718 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.041431 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.051886 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.054324 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.053668 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.070616 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.038481 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.033812 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.045099 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float32, device='cuda')
B = torch.randn(64, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.047778 | 134,217,728 | 4,849,664 | 27.675676 | 4 | {"M": 2048, "N": 512, "K": 64, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.058649 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.025778 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.034601 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.045772 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.031343 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.035114 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.053254 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.043217 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.021611 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.048641 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.054517 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 64) x (64, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 64, dtype=torch.float16, device='cuda')
B = torch.randn(64, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize()
| matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.039634 | 134,217,728 | 2,424,832 | 55.351351 | 2 | {"M": 2048, "N": 512, "K": 64, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.143872 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.079001 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.065847 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.07462 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.088969 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.050464 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.053358 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.083828 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.029083 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.039184 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.06641 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float32, device='cuda')
B = torch.randn(256, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.065831 | 536,870,912 | 6,815,744 | 78.769231 | 4 | {"M": 2048, "N": 512, "K": 256, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.078929 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.036972 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.047303 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.043684 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.030522 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.056294 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.062914 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.051282 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.028663 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.043549 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.041985 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 256) x (256, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 256, dtype=torch.float16, device='cuda')
B = torch.randn(256, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.051707 | 536,870,912 | 3,407,872 | 157.538462 | 2 | {"M": 2048, "N": 512, "K": 256, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.248593 | 1,073,741,824 | 9,437,184 | 113.777778 | 4 | {"M": 2048, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.138667 | 1,073,741,824 | 9,437,184 | 113.777778 | 4 | {"M": 2048, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.085648 | 1,073,741,824 | 9,437,184 | 113.777778 | 4 | {"M": 2048, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.13447 | 1,073,741,824 | 9,437,184 | 113.777778 | 4 | {"M": 2048, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.14091 | 1,073,741,824 | 9,437,184 | 113.777778 | 4 | {"M": 2048, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.089916 | 1,073,741,824 | 9,437,184 | 113.777778 | 4 | {"M": 2048, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.044589 | 1,073,741,824 | 9,437,184 | 113.777778 | 4 | {"M": 2048, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.070859 | 1,073,741,824 | 9,437,184 | 113.777778 | 4 | {"M": 2048, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA RTX 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} | 0.049137 | 1,073,741,824 | 9,437,184 | 113.777778 | 4 | {"M": 2048, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA 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} | 0.06192 | 1,073,741,824 | 9,437,184 | 113.777778 | 4 | {"M": 2048, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA 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} | 0.073185 | 1,073,741,824 | 9,437,184 | 113.777778 | 4 | {"M": 2048, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float32, device='cuda')
B = torch.randn(512, 512, dtype=torch.float32, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA RTX 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} | 0.073306 | 1,073,741,824 | 9,437,184 | 113.777778 | 4 | {"M": 2048, "N": 512, "K": 512, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA T4 | {"gpu_name": "NVIDIA T4", "cuda_cores": 2560, "tensor_cores": 320, "memory_gb": 16, "memory_bandwidth_gbps": 320, "base_clock_mhz": 585, "boost_clock_mhz": 1590, "sm_count": 40, "fp32_tflops": 8.1, "fp16_tflops": 65, "tdp_watts": 70, "compute_capability": 7.5, "l2_cache_mb": 4} | 0.081035 | 1,073,741,824 | 4,718,592 | 227.555556 | 2 | {"M": 2048, "N": 512, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA V100 | {"gpu_name": "NVIDIA V100", "cuda_cores": 5120, "tensor_cores": 640, "memory_gb": 32, "memory_bandwidth_gbps": 900, "base_clock_mhz": 1230, "boost_clock_mhz": 1530, "sm_count": 80, "fp32_tflops": 15.7, "fp16_tflops": 125, "tdp_watts": 300, "compute_capability": 7.0, "l2_cache_mb": 6} | 0.076306 | 1,073,741,824 | 4,718,592 | 227.555556 | 2 | {"M": 2048, "N": 512, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA A10G | {"gpu_name": "NVIDIA A10G", "cuda_cores": 9216, "tensor_cores": 288, "memory_gb": 24, "memory_bandwidth_gbps": 600, "base_clock_mhz": 885, "boost_clock_mhz": 1695, "sm_count": 80, "fp32_tflops": 31.2, "fp16_tflops": 62.5, "tdp_watts": 150, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.05519 | 1,073,741,824 | 4,718,592 | 227.555556 | 2 | {"M": 2048, "N": 512, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA A100 40GB | {"gpu_name": "NVIDIA A100 40GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 40, "memory_bandwidth_gbps": 1555, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.058987 | 1,073,741,824 | 4,718,592 | 227.555556 | 2 | {"M": 2048, "N": 512, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA A100 80GB | {"gpu_name": "NVIDIA A100 80GB", "cuda_cores": 6912, "tensor_cores": 432, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 765, "boost_clock_mhz": 1410, "sm_count": 108, "fp32_tflops": 19.5, "fp16_tflops": 312, "tdp_watts": 400, "compute_capability": 8.0, "l2_cache_mb": 40} | 0.029001 | 1,073,741,824 | 4,718,592 | 227.555556 | 2 | {"M": 2048, "N": 512, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA L4 | {"gpu_name": "NVIDIA L4", "cuda_cores": 7424, "tensor_cores": 232, "memory_gb": 24, "memory_bandwidth_gbps": 300, "base_clock_mhz": 795, "boost_clock_mhz": 2040, "sm_count": 58, "fp32_tflops": 30.3, "fp16_tflops": 121, "tdp_watts": 72, "compute_capability": 8.9, "l2_cache_mb": 48} | 0.076241 | 1,073,741,824 | 4,718,592 | 227.555556 | 2 | {"M": 2048, "N": 512, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA L40S | {"gpu_name": "NVIDIA L40S", "cuda_cores": 18176, "tensor_cores": 568, "memory_gb": 48, "memory_bandwidth_gbps": 864, "base_clock_mhz": 1110, "boost_clock_mhz": 2520, "sm_count": 142, "fp32_tflops": 91.6, "fp16_tflops": 183.2, "tdp_watts": 350, "compute_capability": 8.9, "l2_cache_mb": 96} | 0.04996 | 1,073,741,824 | 4,718,592 | 227.555556 | 2 | {"M": 2048, "N": 512, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA RTX 3090 | {"gpu_name": "NVIDIA RTX 3090", "cuda_cores": 10496, "tensor_cores": 328, "memory_gb": 24, "memory_bandwidth_gbps": 936, "base_clock_mhz": 1395, "boost_clock_mhz": 1695, "sm_count": 82, "fp32_tflops": 35.6, "fp16_tflops": 71, "tdp_watts": 350, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.043962 | 1,073,741,824 | 4,718,592 | 227.555556 | 2 | {"M": 2048, "N": 512, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA RTX 4090 | {"gpu_name": "NVIDIA RTX 4090", "cuda_cores": 16384, "tensor_cores": 512, "memory_gb": 24, "memory_bandwidth_gbps": 1008, "base_clock_mhz": 2235, "boost_clock_mhz": 2520, "sm_count": 128, "fp32_tflops": 82.6, "fp16_tflops": 165.2, "tdp_watts": 450, "compute_capability": 8.9, "l2_cache_mb": 72} | 0.057196 | 1,073,741,824 | 4,718,592 | 227.555556 | 2 | {"M": 2048, "N": 512, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA H100 SXM | {"gpu_name": "NVIDIA H100 SXM", "cuda_cores": 16896, "tensor_cores": 528, "memory_gb": 80, "memory_bandwidth_gbps": 3350, "base_clock_mhz": 1095, "boost_clock_mhz": 1830, "sm_count": 132, "fp32_tflops": 67, "fp16_tflops": 989, "tdp_watts": 700, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.057722 | 1,073,741,824 | 4,718,592 | 227.555556 | 2 | {"M": 2048, "N": 512, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA H100 PCIe | {"gpu_name": "NVIDIA H100 PCIe", "cuda_cores": 14592, "tensor_cores": 456, "memory_gb": 80, "memory_bandwidth_gbps": 2039, "base_clock_mhz": 1095, "boost_clock_mhz": 1620, "sm_count": 114, "fp32_tflops": 48, "fp16_tflops": 756, "tdp_watts": 350, "compute_capability": 9.0, "l2_cache_mb": 50} | 0.034522 | 1,073,741,824 | 4,718,592 | 227.555556 | 2 | {"M": 2048, "N": 512, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 512) x (512, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 512, dtype=torch.float16, device='cuda')
B = torch.randn(512, 512, dtype=torch.float16, device='cuda')
C = matmul_kernel(A, B)
torch.cuda.synchronize... | matmul | NVIDIA RTX A6000 | {"gpu_name": "NVIDIA RTX A6000", "cuda_cores": 10752, "tensor_cores": 336, "memory_gb": 48, "memory_bandwidth_gbps": 768, "base_clock_mhz": 1410, "boost_clock_mhz": 1860, "sm_count": 84, "fp32_tflops": 38.7, "fp16_tflops": 77.4, "tdp_watts": 300, "compute_capability": 8.6, "l2_cache_mb": 6} | 0.054493 | 1,073,741,824 | 4,718,592 | 227.555556 | 2 | {"M": 2048, "N": 512, "K": 512, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 512, 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} | 0.389495 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 512, 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} | 0.194656 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 512, 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} | 0.107246 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 512, 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} | 0.193239 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 512, 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} | 0.179104 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 512, 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} | 0.112121 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 512, 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} | 0.060423 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 512, 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} | 0.134358 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 512, 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} | 0.070765 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 512, 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} | 0.089361 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 512, 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} | 0.081 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float32, device='cuda')
B = torch.randn(1024, 512, 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} | 0.118517 | 2,147,483,648 | 14,680,064 | 146.285714 | 4 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float32"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 512, 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} | 0.093412 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 512, 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} | 0.066911 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 512, 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} | 0.062499 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 512, 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} | 0.052264 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 512, 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} | 0.056823 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 512, 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} | 0.06365 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 512, 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} | 0.067836 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float16"} |
import torch
def matmul_kernel(A, B):
# Matrix multiplication: (2048, 1024) x (1024, 512) -> (2048, 512)
C = torch.matmul(A, B)
return C
A = torch.randn(2048, 1024, dtype=torch.float16, device='cuda')
B = torch.randn(1024, 512, 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} | 0.089043 | 2,147,483,648 | 7,340,032 | 292.571429 | 2 | {"M": 2048, "N": 512, "K": 1024, "dtype": "float16"} |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.