Huawei open-sources neural network stack to challenge CUDA dominance

Huawei commits to releasing its neural network software stack as open source, seeking to boost Artificial Intelligence development around Chinese-designed hardware.

Huawei has announced plans to open-source its Compute Architecture for Neural Networks (CANN) software stack, which powers the company´s Ascend series of processors, by the end of this year. The move, unveiled during Huawei´s developer summit in Shenzhen, is intended to accelerate developer onboarding and cultivate a robust ecosystem around Chinese hardware in the Artificial Intelligence sector. Rotating chairman Eric Xu Zhijun emphasized that open access could galvanize a community-led movement similar to that which propelled Linux and RISC-V into global prominence. Partnering with leading academic institutions, top Artificial Intelligence startups, and the newly established Model-Chips Ecosystem Innovation Alliance, Huawei has crafted transparent contribution and governance frameworks ahead of the public release.

The timing of the announcement is significant, given escalating technical and geopolitical tensions in the Artificial Intelligence hardware landscape. Chinese regulators have launched a security probe into NVIDIA´s H20 graphics processors, citing risks tied to remote management functions and the potential erosion of data sovereignty. Simultaneously, American authorities have granted only limited export permissions for these advanced chips. In stark contrast to NVIDIA´s proprietary CUDA platform—which recently started actively restricting third-party compatibility layers—CANN will welcome forks and encourage cross-platform adaptations from the very start, signaling openness to a more collaborative development model.

While CANN still trails CUDA´s extensive library ecosystem and mature tooling amassed over two decades, Huawei is betting that transparent licensing and active community engagement will help close the gap. If successful, the company could evolve from a hardware manufacturer to a pivotal, neutral orchestrator of a broader open-source Artificial Intelligence framework. The inaugural public version of CANN is expected in the fourth quarter, coinciding with the anticipated debut of the Ascend 910C processors and a reference server platform built in collaboration with Tencent Cloud. By opening up its core technologies, Huawei hopes to catalyze a new era of software innovation, centered not on proprietary lock-in but on cooperative progress in Artificial Intelligence infrastructure.

81

Impact Score

Policymakers weigh pause on Artificial Intelligence data center construction

Federal, state, and local officials are moving to slow or condition large data center development as concerns grow over electricity costs, grid strain, environmental effects, and labor standards. Proposed moratoriums and tax incentive changes are creating new uncertainty for developers, hyperscalers, and financiers.

European Union delays key Artificial Intelligence Act obligations

European Union lawmakers have agreed to revise the Artificial Intelligence Act, delaying major high-risk compliance obligations and easing some overlapping requirements. The changes give businesses more time to prepare while preserving the law’s core framework for high-risk systems and transparency rules.

HMRC signs £175m Quantexa deal for fraud detection

HM Revenue and Customs has signed a £175 million, 10-year agreement with Quantexa to unify fragmented data and strengthen fraud detection. The deployment is designed to automate routine work while keeping decisions transparent, auditable and subject to human approval.

Us supercomputers test new Artificial Intelligence chip suppliers

Sandia National Laboratories is evaluating chips from Israeli startup NextSilicon as major chipmakers shift their roadmaps toward Artificial Intelligence. The move reflects growing concern that mainstream processors are deprioritizing the scientific computing features government labs still need.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.