| | | | |
| Product | NVIDIA® L40S ADA GPU Computing Accelerator - 48GB GDDR6 - PCIe 4.0 x16 - Passive Cooling | NVIDIA® H200 NVL GPU - 141GB HBM3e - PCIe 5.0 x16 - Passive Cooling - 600W | NVIDIA® RTX PRO 4500 Blackwell Server Edition - 32GB GDDR7 ECC - PCIe 5.0 x16 - Passive Cooling | NVIDIA® RTX PRO 6000 Blackwell Server Edition - 96GB GDDR7 - PCIe 5.0 x16 - Passive Cooling (4xDP) - 600W |
| Action | Select | Select | Select | Select |
| Main Specifications |
| Product Series |
Nvidia L40S | Nvidia H200 | | |
| Core Type |
NVIDIA TENSOR | NVIDIA TENSOR | | |
| Host Interface |
PCI Express 4.0 x16 | PCI Express 5.0 x16 | PCI Express 5.0 x16 | PCI Express 5.0 x16 |
| GPU Architecture |
Ada Lovelace | Hopper | | |
| Product Type |
| | Server | Server |
| Product Line |
| | NVIDIA Professional Graphics | NVIDIA Professional Graphics |
| Memory Technology |
| | GDDR7 | GDDR7 |
| Memory Capacity |
| | 32 GB | 96 GB with ECC |
| Max Displays |
| | | 4 Displays |
| Detailed Specifications |
| Streaming Processor Cores |
18,176 | | 10,496 CUDA Parallel Processing Cores | 24,064 CUDA Parallel Processing Cores |
| NVIDIA Tensor Cores |
568 | Gen 4 | | 203 (5th Gen) | 752 (5th Gen) |
| NVIDIA RT Cores |
142 | Gen 3 | | 82 (4th Gen) | 188 (4th Gen) |
| Memory Interface |
| | 256-bit | 512-bit |
| Max Memory Size |
48GB GDDR6 with ECC | 141 GB | | |
| Max Memory Bandwidth |
864 GB/s | 4.8TB/s | | |
| Peak FP64 |
| 30 TFLOPS | | |
| Peak FP64 Tensor Core |
| 60 TFLOPS | | |
| INT8 Tensor Core |
733 teraFLOPS | 3,341 TFLOPS | | |
| TF32 Tensor Core |
183 teraFLOPS | 835 TFLOPS | | |
| FP32 |
91.6 teraFLOPS | 60 TFLOPS | | |
| Peak BFLOAT16 Tensor Core |
362.05 teraFLOPS | 1,671 TFLOPS | | |
| Peak FP16 Tensor Core |
362.05 teraFLOPS | 1,671 TFLOPS | | |
| Peak FP8 Tensor Core |
733 teraFLOPS | 3,341 TFLOPS | | |
| Peak INT4 Tensor Core |
733 teraFLOPS | | | |
| Total NVLink Bandwidth |
Not supported | | | |
| Multi-Instance GPUs |
No | | | |
| NVENC | NVDEC |
3x l 3x (includes AV1 encode and decode) | | | |
| Secure Boot with Root of Trust |
Yes | | | |
| NEBS Ready |
Level 3 | | | |
| Transistor Count |
| | 45.6 Billion | |
| DisplayPort Connectors |
4x DisplayPort 1.4a | | | |
| Cooling |
Passive | Passive | | |
| Dimensions |
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 |
| Dual-slot air-cooled | | |
| Lithography |
| | 4 nm Custom Process | 4N NVIDIA Custom Process |
| Supplementary Power Connectors |
1x 16-pin | | 1x PCIe CEM5 16-pin | 1x PCIe CEM5 16-pin |
| Max Graphics Card Power (W) |
350W | Up to 600W (configurable) | 165W | 400 W – 600 W (configurable) |
| Processor |
| | NVIDIA Blackwell Architecture | NVIDIA Blackwell Architecture |
| Memory Bandwidth |
| | 800 GB/s | Up to 1.6 TB/s |
| Peak Single Precision FP32 Performance |
| | 51 TFLOPS | |
| Minimum Recommended Power, Single Card (W) |
| | | 800 |
| Minimum Recommended Power, 2-Way (W) |
| | | 1600 |
| Minimum Recommended Power, 3-Way (W) |
| | | 2000 |
| Minimum Recommended Power, 4-Way (W) |
| | | 2400 |
| Thermal Solution |
| | Passive | Passive |
| Slot Height |
| | Single Slot | 2-Slot |
| Action | Select | Select | Select | Select |