U.S. restrictions have limited Huawei’s access to advanced chipmaking technology and supply chains, constraining the scale at which it can produce AI chips and helping preserve a gap with Nvidia’s leading systems. But the evidence does not show that Huawei’s chip engineering has stopped advancing. Production estimates conflict, and published performance comparisons lack enough detail to establish a universal benchmark.
What the U.S. restrictions did—and did not do
Two separate actions in May 2025
On May 13, 2025, the Bureau of Industry and Security (BIS) announced that it would rescind the Biden-era AI Diffusion Rule, whose compliance requirements were due to take effect May 15. BIS separately issued guidance about advanced computing chips from China, including Huawei’s Ascend chips. Rescinding the diffusion rule was not the same as removing controls related to Huawei. BIS said it planned to formalize the rescission and issue a replacement rule later.
The Associated Press reported that BIS’s guidance treated Ascend semiconductors as subject to U.S. export controls because the agency believed they used U.S. technology. BIS warned that using them could risk violating U.S. export controls and lead to enforcement action. That is the agency’s stated position; it should not be read as a court-tested conclusion covering every use or transaction.
A later licensing policy addressed different chips
On January 13, 2026, BIS announced case-by-case review of applications to export Nvidia H200, AMD MI325X and similar chips to China, subject to conditions involving supply, purchaser compliance and testing by a U.S. third party. Under Secretary Jeffrey Kessler described the policy this way: “Export controls should evolve with changes in technology, while protecting national security. Permitting the sale of the H200 to China under controlled conditions will strengthen the American technology ecosystem.” This policy concerned applications for those U.S. chips; it did not, by itself, change the treatment of Huawei’s Ascend chips.
#1 Best Overall
- System Compatibility Note: This 2-slot card measures 271 x 112 x 39 mm and requires a single 12V-2x6-pin power connector. Please verify chassis and PSU compatibility before purchase.
- Dedicated Support: Please contact us directly through Amazon for any product questions or assistance you may require.
- Professional Intel Arc Pro B70 GPU: Built on the Intel Xe2-HPG architecture, it features 32 Xe cores and 256 XMX engines, designed to accelerate AI, rendering, and complex visualization workloads.
- Massive 32GB GDDR6 VRAM: Equipped with 32GB of high-speed GDDR6 memory on a 256-bit bus, running at 19 Gbps, which allows for handling large AI models and complex datasets locally.
- High-Performance Engine Clock: Delivers an engine clock of 2540 MHz, providing the compute power needed for demanding professional applications and AI inference.
How many Ascend chips might Huawei produce?
There is no settled production count in the available figures. A 2025 report by the House Select Committee on the Chinese Communist Party describes estimates with different sources and assumptions, ranging from 200,000 to 800,000 chips or equivalents for 2025. They should be treated as estimates, not audited totals.
| Figure | What it represents | Qualification |
|---|---|---|
| No more than 200,000 | Huawei Ascend AI chips manufactured in 2025 | The House Select Committee report attributes this estimate to the U.S. government; it is not a confirmed production count. |
| 250,000 | Equivalent Ascend 910Cs in 2025 | Press reporting cited by the House Select Committee report; not an audited total. |
| As many as 800,000 | Ascend 910Cs in 2025 | An analysis cited by the House Select Committee report, based on a possible high-bandwidth-memory stockpile. |
| Around 300,000 | Ascend 910Cs in 2026 | The same analysis’s conditional projection if high-bandwidth-memory production remained a bottleneck—not a confirmed output figure. |
The report’s range matters: it would be misleading to present one of these estimates as Huawei’s verified annual output. The 2026 figure is especially conditional, since it depends on future memory supply.
Rank #2
- NVIDIA Volta GV100 Architecture — 4,608 CUDA Cores, 640 1st-Gen Tensor Cores delivering 14 TFLOPS FP32 and 112 TFLOPS deep learning performance for AI training, inference, HPC, and scientific computing workloads
- 32GB HBM2 ECC Memory — 900 GB/s Bandwidth — High-bandwidth memory on a 4096-bit bus with ECC error correction provides the memory capacity and throughput required for the largest AI models, simulations, and datasets
- PCIe 3.0 x16 Interface — 250W TDP — Standard PCIe Gen3 connectivity with passive cooling designed for enterprise rack server deployment in HPE ProLiant, Dell PowerEdge, and Supermicro platforms with adequate chassis airflow
- NVLink — Scale to 96GB Unified Memory — Connect two V100 GPUs via NVLink at 300 GB/s bi-directional bandwidth to scale GPU memory from 32GB to 96GB for larger AI training and HPC workloads
- Multi-Precision Computing — Supports FP64 (7 TFLOPS), FP32 (14 TFLOPS), FP16 (112 TFLOPS) and INT8 precision modes for flexible deployment across training, inference, and scientific simulation workloads
What the reported performance gap means
The House Select Committee report characterizes Nvidia’s Blackwell products as substantially faster than Huawei’s Ascend 910C. Its comparisons are:
| Product in the report | Reported comparison with Ascend 910C | How to interpret it |
|---|---|---|
| Nvidia B100 | Roughly 2× the performance | Committee report comparison; the surfaced information does not establish a universal workload or benchmark. |
| Nvidia GB200 | Roughly 3× the performance | Committee report comparison; system configuration and measurement conditions are not sufficiently specified for a general ratio. |
| Nvidia GB300 | Roughly 4× the performance | Committee report comparison; do not treat as a like-for-like result across all uses. |
These ratios are not a substitute for controlled, workload-specific testing. The cited material does not provide a benchmark dataset that establishes parity or a fixed performance multiple across training, inference and different system configurations. Nor does it establish an audited Huawei production disclosure or demonstrate that engineering progress has ceased.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Rank #3
- Professional AI & Creator Workstation: AMD Radeon AI PRO R9700 GPU with 32GB GDDR6 is engineered for AI development, professional content creation, and compute-intensive workloads.
- Massive 32GB Memory Capacity: 32GB of GDDR6 memory on a 256-bit bus provides ample bandwidth for large AI models, 8K video editing, and complex 3D rendering.
- Advanced RDNA 4 with AI Accelerators: 64 Compute Units with 3rd Gen Ray Tracing and dedicated 2nd Gen AI Accelerators for groundbreaking AI performance and visual computing.
- Professional Blower Cooling: Efficient single blower design exhausts heat directly out of the chassis, ideal for multi-GPU workstation and server configurations.
- Enterprise-Grade Thermal Solution: Vapor chamber heatsink with industrial Honeywell PTM7950 thermal interface material ensures reliable cooling under sustained professional loads.
Why chip counts and headline ratios are not the whole story
Restrictions can matter even when they do not halt design work: limited access to advanced manufacturing inputs and supply chains can constrain how many accelerators are made and deployed. High-bandwidth memory (HBM) is one reported bottleneck in the production estimates. For a buyer, policymaker or investor assessing the practical significance, available volume and the ability to combine chips into usable systems matter alongside a chip’s nominal speed.
A meaningful comparison between AI accelerators needs to specify the conditions being compared:
Rank #4
- System Compatibility Note: 2-slot card, 271x112x39mm, single 8-pin power, 200W TDP. Verify chassis clearance and PSU capacity before purchase.
- Dedicated Support: Please contact us directly through Amazon for any product questions or assistance you may require.
- 24GB GDDR6 on 192-Bit Bus: Massive 24GB memory with 456 GB/s bandwidth – ideal for LLMs, AI inference, 3D rendering, and generative design.
- Intel Xe2-HPG Architecture: Built on Intel's next-gen architecture with 20 Xe cores and 160 XMX engines for AI acceleration (197 INT8 TOPS).
- PCIe 5.0 Support: PCI Express 5.0 x16 interface for maximum bandwidth with the latest workstation platforms.
- Workload: model training and inference can produce different performance results.
- Model and precision: the model being run and numerical precision affect both speed and memory demand.
- Measurement scale: single-chip results are not interchangeable with whole-system or cluster results.
- Memory: capacity, bandwidth and access to HBM can affect whether a system can run a workload at all.
- Interconnect and scaling: communication between chips matters when a workload spans multiple accelerators.
- Software: compatibility and optimization affect how much of a chip’s theoretical capability is usable.
- Availability: manufacturing volume and deployment scale affect whether systems can be obtained and put to work.
The committee report’s performance ratios do not specify enough of these conditions to resolve a like-for-like comparison. They are useful as a report-level indication of a claimed gap, not as a universal measure of real-world performance.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Is Huawei’s AI-chip technology literally stuck in the past?
No. “In the past” is a forceful way to describe the pressure restrictions and bottlenecks can put on manufacturing scale and competitiveness, but the figures do not prove that Huawei’s technology has stopped advancing. They show contested production estimates, a reported memory constraint and a committee-level comparison with Nvidia products whose benchmark conditions are not fully established. Any present-day conclusion about export rules also needs to account for policy changes after the dated actions described above.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Clear out junk files and repair common Windows errors3Fix the driver behind crashes, sound loss and screen glitchesQuick Recap
Best Value
- PLEASE NOTE: Exporting an NVIDIA RTX Pro 6000 GPU outside the US requires strict adherence to the U.S. Export Administration Regulations (EAR) and issuance of an export license from the Bureau of Industry and Security (BIS). Compliance and Know Your Customer (KYC) screening may be required as a condition of order acceptance. [NVIDIA Blackwell Streaming Multiprocessor] The new SM features increased processing throughput, and new neural shaders that integrate neural networks inside of programmable shaders | DLSS 4: Multi Frame Generation ensures ultra-smooth frame pacing for lifelike simulations.
- [Double-Flow-Through Design] The RTX PRO 6000 Blackwell features a double-flow-through cooling design, optimizing efficiency and airflow to sustain peak performance under 600W power loads. | [5th Gen Tensor Cores] Deliver up to 3X the performance of the previous generation and support for FP4 precision for faster AI model processing times with reduced memory usage, enabling local fine-tuning of LLMs and generative AI | [4th Gen Ray Tracing Cores] Double the ray-triangle intersection rate of the previous generation to create photoreal, physically accurate scenes and immersive 3D designs with RTX Mega Geometry, which enables up to 100X more ray-traced triangles.
- [PCIe Gen 5] Support for PCIe Gen 5 provides double the bandwidth of PCIe Gen 4, improving data-transfer speeds from CPU memory and unlocking faster performance for data-intensive tasks like AI, data science, and 3D modeling. | [GDDR7 Memory] With 96 GB of GPU memory and 1.8 TB ps bandwidth, it can tackle massive 3D and AI projects, fine-tune AI models locally, explore large-scale VR environments, and drive larger multi-app workflows.
- [DisplayPort 2.1] Achieve unparalleled visual clarity and performance, driving high resolution displays at up to 8K at 240 Hz and 16K at 60 Hz. Increased bandwidth enables seamless multi-monitor setups while HDR and higher color depth support ensures superior color accuracy for precision work, such as video editing, 3D design, and live broadcasting.
- [Universal MIG] Divide a single RTX PRO 6000 Blackwell into multiple isolated instances, each with dedicated resources, allowing for concurrent execution of multiple workloads, optimized GPU utilization, and secure isolation of different applications or users. [WARRANTY] 3 YR Manufacturer's Warranty. Bulk OEM Packaging. Retail Packaging is NOT included.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
