Roadmap chip — Huawei Connect 2025 announcement, not yet shipping
HU
Huawei Ascend 970
Da Vinci v5 OAM 2028
VRAM
288 GB
Bandwidth
14.4 TB/s
memory
Performance Metrics
Peak theoretical throughput by precision type
| Precision | Bits | Peak TFLOPS | |
|---|---|---|---|
| FP8 | 8 | 4000.0 | |
| FP4 | 4 | 8000.0 | |
| FP16 | 16 | 2000.0 | |
| BF16 | 16 | 2000.0 |
Power Specifications
TDP
--
Max Power
--
Power Connector
PCIe Slot
Cooling
Air
Memory Specifications
Capacity
288 GB
Type
--
Bandwidth
14400 GB/s
Interface
--
Hardware & Design
Form Factor
OAM
Architecture
Da Vinci v5
Process Node
--
Launch Year
2028
Variant
Standard
Market Segment
Data Center
Full Specifications
| Memory | |
|---|---|
| VRAM | 288 GB |
| Bandwidth | 14.4 TB/s |
| Interconnect & I/O | |
| GPU-to-GPU | Unified Bus 3.0 |
| Interconnect Bandwidth | 4.0 TB/s |
| Power & Thermal | |
| General | |
| Form Factor | OAM |
| Architecture | Da Vinci v5 |
| Launch Year | 2028 |
Documentation & Resources
Common Use Cases
General Compute AI/ML Workloads Data Processing
The Huawei Ascend 970 is optimized for high-performance computing tasks with Da Vinci v5 architecture delivering high TFLOPS of compute power.
Where to Rent
Compare cloud providers offering on-demand GPU instances for AI training, inference, and HPC workloads.
Browse GPU Cloud ProvidersStay Updated on GPU Releases
Get notified when new GPUs are added or specifications are updated.
Loading verification...
No spam, unsubscribe anytime.