NVIDIA A2 - 16GB GDDR6 Single-Slot Low-Profile PCIe 4.0 x8 GPUSKU: GPUNTA2 | |
| # CUDA Cores: | 1280 |
| # RT Cores: | 10 - Gen 2 |
| # Tensor Cores: | 40 - Gen 3 |
| Game Clock: | 1440 MHz Base - 1770 MHz Boost |
| Manufacturer: | PNY |
| Memory: | 16 GB GDDR6 ECC |
| Peak FP32 Performance: | 4.5 TFLOPS |
| Peak FP32 Tensor Core: | 18 TFLOPS - Sparsity |
| Peak Int4 Performance: | 144 TOPS - Sparsity |
| Peak Int8 Performance: | 72 TOPS - Sparsity |
NVIDIA A2 - 16GB GDDR6 Single-Slot Low-Profile PCIe 4.0 x8 GPUSKU: GPUNTA2 | |
| # CUDA Cores: | 1280 |
| # RT Cores: | 10 - Gen 2 |
| # Tensor Cores: | 40 - Gen 3 |
| Game Clock: | 1440 MHz Base - 1770 MHz Boost |
| Manufacturer: | PNY |
| Memory: | 16 GB GDDR6 ECC |
| Peak FP32 Performance: | 4.5 TFLOPS |
| Peak FP32 Tensor Core: | 18 TFLOPS - Sparsity |
| Peak Int4 Performance: | 144 TOPS - Sparsity |
| Peak Int8 Performance: | 72 TOPS - Sparsity |
NVIDIA L4 - 24GB Single Slot, Low Profile PCIe Tensor Core GPU CardSKU: GPUNTL4 | |
| GPU Clock: | 795 MHz Base - 2040 MHz Boost |
| Manufacturer: | PNY |
| Peak BFLOAT16 Tensor Core: | 242 TFLOPS - Sparsity |
| Peak FP16 Tensor Core: | 242 TFLOPS - Sparsity |
| Peak FP32: | 30.3 |
| Peak FP8 Tensor Core: | 485 TFLOPS - Sparsity |
| Peak INT8 Tensor Core: | 485 TOPS - Sparsity |
NVIDIA L4 - 24GB Single Slot, Low Profile PCIe Tensor Core GPU CardSKU: GPUNTL4 | |
| GPU Clock: | 795 MHz Base - 2040 MHz Boost |
| Manufacturer: | PNY |
| Peak BFLOAT16 Tensor Core: | 242 TFLOPS - Sparsity |
| Peak FP16 Tensor Core: | 242 TFLOPS - Sparsity |
| Peak FP32: | 30.3 |
| Peak FP8 Tensor Core: | 485 TFLOPS - Sparsity |
| Peak INT8 Tensor Core: | 485 TOPS - Sparsity |
NVIDIA A10 Passive - 24GB GDDR6 PCIe 4.0 150W GPU ModuleSKU: GPUNA1024GB |
NVIDIA A16 Passive - 4x 16GB ECC GDDR6 PCIe 4.0 250W GPU ModuleSKU: GPUNA164x16GB |
NVIDIA A30 Passive - 24GB ECC HBM2 PCIe 4.0 165W GPU ModuleSKU: GPUNA3024GB |
NVIDIA A30X Passive - 24GB ECC HBM2 PCIe 4.0 165W Converged GPU Accelerator ModuleSKU: GPUNA30X24GB | |
| # Transistors: | 54.2 Billion |
| Architecture: | Ampere |
| GPU Memory: | 24 GB HBM2e |
| Networking: | 2x 100 Gbps ports, Ethernet or InfiniBand |
| Peak FP16 Tensor Core: | 165 TFLOPS - Sparsity |
| Peak FP32: | 10.3 TFLOPS |
| Peak FP64: | 5.2 TFLOPS |
| Peak FP64 Tensor Core: | 10.3 TFLOPS - Sparsity |
| Peak INT8 Tensor Core: | 330 TOPS - Sparsity |
| Peak TF32 Tensor Core: | 82.6 TFLOPS - Sparsity |
NVIDIA A30X Passive - 24GB ECC HBM2 PCIe 4.0 165W Converged GPU Accelerator ModuleSKU: GPUNA30X24GB | |
| # Transistors: | 54.2 Billion |
| Architecture: | Ampere |
| GPU Memory: | 24 GB HBM2e |
| Networking: | 2x 100 Gbps ports, Ethernet or InfiniBand |
| Peak FP16 Tensor Core: | 165 TFLOPS - Sparsity |
| Peak FP32: | 10.3 TFLOPS |
| Peak FP64: | 5.2 TFLOPS |
| Peak FP64 Tensor Core: | 10.3 TFLOPS - Sparsity |
| Peak INT8 Tensor Core: | 330 TOPS - Sparsity |
| Peak TF32 Tensor Core: | 82.6 TFLOPS - Sparsity |
NVIDIA L40 - 48GB GDDR6 ECC PCIe GPU CardSKU: GPUNTL40 | |
| # CUDA Cores: | 18176 |
| Interfaces: | 4x DP 1.4a |
| Manufacturer: | PNY |
| Memory: | 48GB GDDR6 ECC |
| # RT Cores: | 142 - Gen 3 |
| # Tensor Cores: | 568 - Gen 4 |
NVIDIA L40 - 48GB GDDR6 ECC PCIe GPU CardSKU: GPUNTL40 | |
| # CUDA Cores: | 18176 |
| Interfaces: | 4x DP 1.4a |
| Manufacturer: | PNY |
| Memory: | 48GB GDDR6 ECC |
| # RT Cores: | 142 - Gen 3 |
| # Tensor Cores: | 568 - Gen 4 |
NVIDIA A800 40GB HBM2 Active GPUSKU: GPUNA80040GB |
NVIDIA A100 Passive - 80GB ECC HBM2 PCIe 4.0 300W GPU Module - 312TFLOPs DL / 19.5TFLOPs SP / 9.7TFLOPs DPSKU: GPUNTA10080GBPCIE | |
| # CUDA Cores: | 6912 |
| # Streaming Multiprocessors: | 108 |
| # Transistors: | 54 Billion |
| Manufacturer: | PNY |
| Memory: | 80GB HBM2e ECC on by Default |
| Peak FP16 Tensor Core: | 624 TFLOPS - Sparcity |
| Peak FP32 Tensor Core: | 312 TFLOPS - Sparcity |
| Peak FP64: | 5.2 TFLOPS |
| Peak FP64 Tensor Core: | 19.5 TFLOPS |
| Peak INT8 Tensor Core: | 1248 TOPS - Sparcity |
| Peak FP32: | 19.2 TFLOPS |
NVIDIA A100 Passive - 80GB ECC HBM2 PCIe 4.0 300W GPU Module - 312TFLOPs DL / 19.5TFLOPs SP / 9.7TFLOPs DPSKU: GPUNTA10080GBPCIE | |
| # CUDA Cores: | 6912 |
| # Streaming Multiprocessors: | 108 |
| # Transistors: | 54 Billion |
| Manufacturer: | PNY |
| Memory: | 80GB HBM2e ECC on by Default |
| Peak FP16 Tensor Core: | 624 TFLOPS - Sparcity |
| Peak FP32 Tensor Core: | 312 TFLOPS - Sparcity |
| Peak FP64: | 5.2 TFLOPS |
| Peak FP64 Tensor Core: | 19.5 TFLOPS |
| Peak INT8 Tensor Core: | 1248 TOPS - Sparcity |
| Peak FP32: | 19.2 TFLOPS |
NVIDIA A100X Passive - 80GB ECC HBM2 PCIe 4.0 300W Converged GPU Accelerator Module - 312TFLOPs DL / 19.5TFLOPs SP / 9.7TFLOPs DPSKU: GPUNTA100X80GBPCIE | |
| Manufacturer: | PNY |
| Memory: | 80GB HBM2e |
| Peak FP16 Tensor Core: | 318.5 TFLOPS - Sparsity |
| Peak FP32: | 19.9 |
| Peak FP32 Tensor Core: | 159 TFLOPS - Sparsity |
| Peak FP64: | 9.9 |
| Peak FP64 Tensor Core: | 19.1 TFLOPS - Sparsity |
| Peak INT8 Tensor Core: | 637 TOPS - Sparsity |
NVIDIA A100X Passive - 80GB ECC HBM2 PCIe 4.0 300W Converged GPU Accelerator Module - 312TFLOPs DL / 19.5TFLOPs SP / 9.7TFLOPs DPSKU: GPUNTA100X80GBPCIE | |
| Manufacturer: | PNY |
| Memory: | 80GB HBM2e |
| Peak FP16 Tensor Core: | 318.5 TFLOPS - Sparsity |
| Peak FP32: | 19.9 |
| Peak FP32 Tensor Core: | 159 TFLOPS - Sparsity |
| Peak FP64: | 9.9 |
| Peak FP64 Tensor Core: | 19.1 TFLOPS - Sparsity |
| Peak INT8 Tensor Core: | 637 TOPS - Sparsity |
