He claimed that this dynamic leaves alternative architectures — such as domestic AI chips incompatible with CUDA — with almost no route to widespread adoption, regardless of theoretical performance. In his view, if China becomes similarly locked into this ecosystem, it will effectively lose sovereignty over its AI trajectory.
Wei is clearly not alone in this line of thought. Chinese companies such as Cambricon and Huawei have advanced CUDA alternatives and software abstraction layers in parallel with domestic AI hardware, with state-aligned investment efforts backing broader ecosystem development. Cambricon’s NeuWare stack, for instance, now offers compatibility with PyTorch, TensorFlow, and ONNX, and provides migration tools for CUDA-trained models. Alibaba and Huawei have similarly promoted their own developer frameworks targeting Ascend and XuanTie hardware.
Wei warned that continued reliance on U.S.-controlled software and toolchains would carry “very serious” consequences, both economically and geopolitically. He said China’s goal is to “abandon U.S. technology routes” and build a domestic AI technology stack that is resilient to supply chain shocks and regulatory constraints.
The Chinese are right to be concerned. In recent years, the U.S. has significantly expanded its export controls and blocked Nvidia from selling its leading hardware to Chinese firms. In response, the company introduced modified, lower-performance chips , which were also eventually restricted. Today, now that Nvidia has no market share left in China , domestic companies rely on legacy hardware or workarounds using second-tier accelerators and the grey market.
The claims made at ICC raise questions that cannot yet be answered without detailed benchmarks and independent testing. While hybrid bonding is an established technology, the real-world performance of a full stack built on 14nm logic and 18nm DRAM remains to be seen.
Thermal dissipation in 3D stacks is still a major concern, particularly when logic and memory are in close proximity, and airflow is limited. Hybrid bonding also requires very high manufacturing precision and wafer alignment. These challenges could impact both yield and cost. China’s domestic foundries, including SMIC, are known to have strong 14nm-class production lines, but hybrid bonding at scale for logic-memory stacks would be a new capability.
Breaking Nvidia’s lock on the model training ecosystem also requires not just raw performance parity, but broad software support and capable developers. Even if China succeeds in building a functionally equivalent AI processor, adoption will remain limited without tooling and integration into PyTorch or TensorFlow workflows.
Still, with EUV lithography out of reach and GAA transistor designs still years away for domestic fabs, architectural innovation and packaging integration may be China’s most viable path to short-term competitiveness in high-performance AI workloads. Wei indicated that more technical details would be shared in future disclosures, but did not confirm whether working silicon had been produced.
Follow Tom's Hardware on Google News , or add us as a preferred source , to get our latest news, analysis, & reviews in your feeds.
Luke James Social Links Navigation Contributor Luke James is a freelance writer and journalist. Although his background is in legal, he has a personal interest in all things tech, especially hardware and microelectronics, and anything regulatory.
Key considerations
- Investor positioning can change fast
- Volatility remains possible near catalysts
- Macro rates and liquidity can dominate flows
Reference reading
- https://www.tomshardware.com/tech-industry/semiconductors/SPONSORED_LINK_URL
- https://www.tomshardware.com/tech-industry/semiconductors/china-claims-14nm-ai-chip-can-rival-nvidia-4nm-gpus#main
- https://www.tomshardware.com
- Chinese startup founded by Google engineer claims to have developed its own TPU chip for AI — custom ASIC reportedly 1.5 times faster than Nvidia's A100 GPU fro
- China's hybrid-bonded AI accelerators could rival Nvidia's Blackwell GPUs — top semiconductor expert hints at 'fully controllable domestic solution'
- Grab the Meta Quest S3 at its lowest ever price for Black Friday — just £229.99 for this standalone 120Hz VR headset with 128GB storage and a free game
- DOOM gets ported to board design app, transforming walls into PCB traces, iconic demons into 64-pin packages, and ammo into 3-pin parts — fully playable KiCAD e
- Powering AI Superfactories, NVIDIA and Microsoft Integrate Latest Technologies for Inference, Cybersecurity, Physical AI
Informational only. No financial advice. Do your own research.