| | | | | | | | |
Product | AMD Instinct™ MI210 Accelerator - 64GB HBM2e - PCIe 4.0 x16 - Passive Cooling | NVIDIA® L4 ADA GPU Computing Accelerator - 24GB GDDR6X - PCIe 4.0 x16 - Passive Cooling | NVIDIA® L40S ADA GPU Computing Accelerator - 48GB GDDR6 - PCIe 4.0 x16 - Passive Cooling | NVIDIA® RTX 4000 Ada Generation - 20GB GDDR6 ECC - PCIe 4.0 x16 - Active Cooling (4xDP) | NVIDIA® RTX 4500 Ada Generation - 24GB GDDR6 ECC - PCIe 4.0 x16 - Active Cooling (4xDP) | NVIDIA® RTX 5000 Ada Generation - 32GB GDDR6 ECC - PCIe 4.0 x16 - Active Cooling (4xDP) | NVIDIA® RTX 6000 Ada Generation - 48GB GDDR6 ECC - PCIe 4.0 x16 - Active Cooling (4xDP) | NVIDIA® RTX PRO 6000 Blackwell Max-Q Workstation Edition - 96GB GDDR7 ECC - PCIe 5.0 x16 - Active Cooling (4xDP) |
Action | Select | Select | Select | Select | Select | Select | Select | Select |
Main Specifications |
Product Series |
AMD Instinct | Nvidia L4 | Nvidia L40S | | | | | |
Core Type |
| NVIDIA TENSOR | NVIDIA TENSOR | | | | | |
Core Clock Speed |
1700 MHz | 795 MHz Base | 2040 MHz Boost | | | | | | |
Host Interface |
PCI Express 4.0 x16 | PCI Express 4.0 x16 | PCI Express 4.0 x16 | PCI Express 4.0 x16 | PCI Express 4.0 x16 | PCI Express 4.0 x16 | PCI Express 4.0 x16 | PCI Express 5.0 x16 |
GPU Architecture |
CDNA2 | Ada Lovelace | Ada Lovelace | | | | | |
Product Type |
| | | | Workstation | Workstation | Workstation | Workstation |
Product Line |
| | | NVIDIA Professional Graphics | NVIDIA Professional Graphics | NVIDIA Professional Graphics | NVIDIA Professional Graphics | NVIDIA Professional Graphics |
Memory Technology |
| | | GDDR6 | GDDR6 | GDDR6 | GDDR6 | GDDR7 |
Memory Capacity |
| | | 20 GB GDDR6 ECC | 24 GB GDDR6 ECC | 32 GB GDDR6 ECC | 48 GB with ECC | 96 GB with ECC |
Max Displays |
| | | | | 4 Displays | 4 Displays | 4 Displays |
Detailed Specifications |
Streaming Processor Cores |
6656 | | 18,176 | 6144 CUDA Cores | 7,680 CUDA Parallel Processing Cores | 12,800 CUDA Parallel Processing Cores | 18,176 | 24,064 CUDA Parallel Processing Cores |
Compute Units |
104 | | | | | | | |
NVIDIA Tensor Cores |
| | 568 | Gen 4 | 192 | 240 | 400 | 568 | 752 |
NVIDIA RT Cores |
| | 142 | Gen 3 | 48 | 60 | 100 | 142 | 188 |
Memory Clock Speed |
1.6 GHz | 6251 MHz | | | | | | |
Memory Interface |
4096-bit | 192-bit | | 160-bit | 192-bit | 256-bit | 384-bit | 512-bit |
Max Memory Size |
64 GB HBM2e | 24 GB | 48GB GDDR6 with ECC | | | | | |
Max Memory Bandwidth |
Up to 1638.4 GB/s | 300 GB/s | 864 GB/s | | | | | |
ECC Protection |
Yes (Full-Chip) | On by Default | | | | | | |
Infinity Fabric™ Links |
3 | | | | | | | |
Peak Infinity Fabric™ Link Bandwidth |
100 GB/s | | | | | | | |
INT8 Tensor Core |
| 485 TOPS | Sparsity | 733 teraFLOPS | | | | | |
TF32 Tensor Core |
| 120 TFLOPS | Sparsity | 183 teraFLOPS | | | | | |
FP32 |
22.6 TFLOPs | 30.3 TFLOPS | 91.6 teraFLOPS | | | | | |
Peak BFLOAT16 Tensor Core |
| 242 TFLOPS | Sparsity | 362.05 teraFLOPS | | | | | |
Peak FP16 Tensor Core |
| 242 TFLOPS | Sparsity | 362.05 teraFLOPS | | | | | |
Peak FP8 Tensor Core |
| 485 TFLOPS | Sparsity | 733 teraFLOPS | | | | | |
Peak INT4 Tensor Core |
| | 733 teraFLOPS | | | | | |
Total NVLink Bandwidth |
| | Not supported | | | | | |
Multi-Instance GPUs |
| | No | | | | | |
Tensor Performance |
| | | 327.6 TFLOPS | 637.8 TFLOPS | 1044.4 TFLOPS | 1457.0 TFLOPS | |
NVENC | NVDEC |
| 2 | 4 | 4 | JPEG Decoders | AV1 Encode and Decode | 3x l 3x (includes AV1 encode and decode) | | | | | |
Secure Boot with Root of Trust |
| Yes | Yes | | | | | |
NEBS Ready |
| Yes | Level 3 | Level 3 | | | | | |
Peak Single Precision Matrix (FP32) Performance |
45.3 TFLOPs | | | | | | | |
Peak Double Precision Matrix (FP64) Performance |
45.3 TFLOPs | | | | | | | |
Peak Double Precision (FP64) Performance |
22.6 TFLOPs | | | | | | | |
Peak INT4 Performance |
181 TOPs | | | | | | | |
Peak bfloat16 |
181 TFLOPs | | | | | | | |
Transistor Count |
| | | 35.8 Billion | 35.8 billion | 76.3 billion | 76.3 billion | 92.2 Billion |
DisplayPort Connectors |
| None | vGPU Only | 4x DisplayPort 1.4a | | | | | |
OS Support |
Linux x86_64 | | | | | | | |
Cooling |
Passive | Passive | Passive | | | | | |
Dual Slot |
yes | No | | | | | | |
Dimensions |
10.5" (267 mm) Board Length | | 4.4" (H) x 10.5" (L) | 4.4" (H) x 9.5"(L) | 4.4" H x 10.5" L | 4.4" H x 10.5" L | 4.4" H x 10.5" L | 4.4” H x 10.5” L, FHFL Dual Slot |
Form Factor |
Full Height | 6.61” L x 2.71” H (Low-profile) | | | | | | |
Lithography |
| | | | | 4 nm NVIDIA Custom Process | 4 nm NVIDIA Custom Process | 4N NVIDIA Custom Process |
Supplementary Power Connectors |
1x8 pin 12V EPS | | 1x 16-pin | | 1x 16-pin CEM5 PCIe | 1x 16-pin CEM5 PCIe | 1x PCIe CEM5 16-pin | 1x PCIe CEM5 16-pin |
Max Graphics Card Power (W) |
300W Peak | 72W | 350W | 130W | 210W | 250W | 300W | 300W |
Processor |
| | | NVIDIA Ada Lovelace | NVIDIA Ada Lovelace | NVIDIA Ada Lovelace | NVIDIA Ada Lovelace | NVIDIA Blackwell Architecture |
Memory Bandwidth |
| | | 360 GB/s | 432 GB/s | 576 GB/s | 960 GB/s | 1792 GB/s |
Peak Single-Precision Performance |
| | | | | 65.3 TFLOPS | | |
Peak Single Precision FP32 Performance |
| | | 26.7 TFLOPS | 39.9 TFLOPS | | 91.1 TFLOPS | |
NVLink Interconnect |
| | | Not Supported | Not Supported | Not Supported | | |
RT Core Performance |
| | | 61.8 TFLOPS | 92.2 TFLOPS | 151.0 TFLOPS | 210.6 TFLOPS | |
DisplayPort Output |
| | | | | | 4x DP 1.4a | 4x DP 2.1 |
Mini DisplayPort Output |
| | | 4x mDP 1.4a | 4x DP 1.4a | 4x DP 1.4a | | |
Minimum Recommended Power, Single Card (W) |
| | | | | | 600 | 600 |
Minimum Recommended Power, 2-Way (W) |
| | | | | | 750 | 900 |
Minimum Recommended Power, 3-Way (W) |
| | | | | | 850 | 1200 |
Minimum Recommended Power, 4-Way (W) |
| | | | | | 1000 | 1600 |
Thermal Solution |
| | | Blower Active Fan | Blower Active Fan | Blower Active Fan | Blower Active Fan | Blower Active Fan |
Slot Height |
| | | Single Slot | 2-Slot | 2-Slot | 2-Slot | 2-Slot |
Action | Select | Select | Select | Select | Select | Select | Select | Select |