Flopper.io GPU Comparison Sheet
NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition vs NVIDIA GB200 Grace Blackwell Superchip 372GB
Side by side datacenter GPU specifications from the Flopper.io database.
Overview
| NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition | NVIDIA GB200 Grace Blackwell Superchip 372GB | |
|---|---|---|
| FP32 TFLOPS | 109.7 | 160.0 |
| VRAM | 96 GB | 372 GB |
| Memory Bandwidth | 1.8 TB/s | 16.0 TB/s |
| TDP | 300 W | — |
Performance by Precision
| Precision | NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition peak TFLOPS (dense) | NVIDIA GB200 Grace Blackwell Superchip 372GB peak TFLOPS (dense) |
|---|---|---|
| FP64 | 1.7 | 80.0 |
| FP32 | 109.7 | 160.0 |
| TF32 | 219.5 | 2500.0 |
| BF16 | 438.9 | 5000.0 |
| FP16 | 438.9 | 5000.0 |
| FP8 | 877.9 | 10000.0 |
| FP6 | — | 10000.0 |
| FP4 | 1755.7 | 20000.0 |
| INT8 | 877.9 | 10000.0 |
Specifications
| NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition | NVIDIA GB200 Grace Blackwell Superchip 372GB | |
|---|---|---|
| Architecture | Blackwell | Blackwell |
| Process Node | 4nm | 4nm |
| Launch Year | 2025 | 2024 |
| Form Factor | PCIe | Superchip |
| Memory Type | GDDR7 | HBM3e |
| Interface Width | 512-bit | — |
| Interconnect | — | NVLink-C2C |
| TDP | 300 W | — |
| FP32 Efficiency | 0.37 TFLOPS/W | — |