Specifications
GPU Architecture
NVIDIA Blackwell
CUDA Cores
10,496
GPU Memory
32GB GDDR7
Memory Interface
256-bit
Memory Bandwidth
800 GB/s
FP4 Tensor Core
1.6 PFLOPS
FP8 Tensor Core
811 TFLOPS
FP16/BF16
406 TFLOPS
System Interface
PCIe 5.0 x16
Power
165 W
Form Factor
4.4" H x 10.5" L, single slot, passive
Multi-Instance GPU
Up to 2 MIG @ 16GB
Full Specifications
| Parameter | Value |
|---|---|
| GPU Architecture | NVIDIA Blackwell |
| CUDA Cores | 10,496 |
| GPU Memory | 32GB GDDR7 |
| Memory Interface | 256-bit |
| Memory Bandwidth | 800 GB/s |
| FP4 Tensor Core | 1.6 PFLOPS |
| FP8 Tensor Core | 811 TFLOPS |
| FP16/BF16 | 406 TFLOPS |
| System Interface | PCIe 5.0 x16 |
| Power | 165 W |
| Form Factor | 4.4" H x 10.5" L, single slot, passive |
| Multi-Instance GPU | Up to 2 MIG @ 16GB |
Application
Passive, single-slot server GPU optimized for dense multi-GPU AI inference, vision AI and video content understanding in data center and edge servers.
Request a Quote — NVIDIA RTX PRO 4500 Blackwell Server Edition 900-5G147-0150-000
QS Compute — AI computing hardware supply partner. Global B2B logistics, competitive pricing, full manufacturer warranty.
Get Your Quote →