Flopper.io GPU Comparison Sheet
Huawei Atlas 350 vs NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition
Side by side datacenter GPU specifications from the Flopper.io database.
Overview
| Huawei Atlas 350 | NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition | |
|---|---|---|
| FP32 TFLOPS | — | 109.7 |
| Memory | 112 GB | 96 GB |
| Memory Bandwidth | 1.4 TB/s | 1.8 TB/s |
| TDP | 600 W | 300 W |
Performance by Precision
| Precision | Huawei Atlas 350 peak TFLOPS (dense) | NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition peak TFLOPS (dense) |
|---|---|---|
| FP64 | — | 1.7 |
| FP32 | — | 109.7 |
| TF32 | — | 219.5 |
| BF16 | 425.0 | 438.9 |
| FP16 | 425.0 | 438.9 |
| FP8 | 804.0 | 877.9 |
| FP6 | 804.0 | — |
| FP4 | 1561.0 | 1755.7 |
| INT8 | 804.0 | 877.9 |
Specifications
| Huawei Atlas 350 | NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition | |
|---|---|---|
| Architecture | Da Vinci v3 | Blackwell |
| Process Node | — | 4nm |
| Launch Year | 2026 | 2025 |
| Form Factor | Custom | PCIe |
| Memory Type | HiBL 1.0 | GDDR7 |
| Interface Width | — | 512-bit |
| Interconnect | Unified Bus 2.0 | — |
| TDP | 600 W | 300 W |
| FP32 Efficiency | — | 0.37 TFLOPS/W |