Chips & InfrastructureSeptember 30, 2026

DeepSeek open-sources tools for Huawei’s AI chips

Its engineers added Huawei’s Ascend 950 to TileLang, an open-source language for AI chip code. DeepSeek says Huawei gave full support.

DeepSeek, the Hangzhou-based AI developer, announced the release on its official WeChat account on September 30. It came two weeks after Huawei unveiled its next generation of AI chips and computing systems.

By The Open Source ScoutOPEN SOURCE & CHIPS

Published: September 30, 2026, 09:41 CEST
Reading time: 3 min
Evidence: 3 sources · 1 primary

This story was generated by AI systems under the editorial rules of Silicon AI News and checked against the sources it links. Found a mistake? Email editor@siliconainews.com. How we report

In brief

DeepSeek said on September 30 that it is open-sourcing programming tools for Huawei’s Ascend AI chips, including compute and communication libraries. Its engineers built support for the Ascend 950 into TileLang, an open-source language for writing the code that runs on AI chips. Reuters and the South China Morning Post describe it as a step away from Nvidia’s software.

New to this? Read it in simple words

DeepSeek is a Chinese AI company that shares much of its work for free.
It released free tools that help programs run on Huawei’s AI chips.
Most AI programs today are written for Nvidia’s chips and software.
The new tools could make switching easier, but nobody outside has tested them.

Words to know

Open source: Software whose code anyone can read, use and change.
Kernel: A small program that does one heavy maths job on a chip.
CUDA: Nvidia’s software that most AI programs use to run on its chips.

What DeepSeek released

DeepSeek released six software modules for Huawei’s Ascend platform, the South China Morning Post reports. They mirror tools DeepSeek had already open-sourced for Nvidia’s chips.

Among them are compute and communication libraries. DeepSeek said Huawei gave “full support”, and that the two companies jointly advanced a “supernode” design built on 128 Ascend 950 chips, Reuters reports.

Sources2 3

KERNEL CODE IN TILELANG→NVIDIA GPUS · CUDA (MAIN)+HUAWEI ASCEND 950 · NEW

BUILT BY DEEPSEEK ENGINEERS · HUAWEI SUPPORT · SPEED NOT YET TESTED OUTSIDE

SOFTWARE STACK 01

One language, two kinds of chips.

From TileLang’s project pages and Reuters. Speed on Huawei’s chip is not yet independently tested.

The role of TileLang

TileLang is an open-source language for writing kernels, the small programs that do the heavy maths on AI chips. It was first released in January 2025 by developers supervised at Peking University, and Nvidia’s CUDA is its main target.

On September 30, the project added official support for Huawei’s Ascend 950. Its notes say engineers from DeepSeek mainly developed this part, and thank “the Huawei team for their close collaboration and valuable support”.

DeepSeek says TileLang offers “a simpler programming model” than CUDA, Nvidia’s software platform.

Sources1 2 3

What is not known yet

The project publishes its own speed charts for the Ascend 950, compared with Huawei’s PyTorch tools. No outside group has tested them yet, and DeepSeek did not announce a new model tuned for Huawei’s chips.

DeepSeek said the first step towards an “independent, self-controlled” software ecosystem for AI chips is a language that is universal, easy to program and still reaches the hardware’s full performance.

Sources1 2

Sources

Every fact in this story comes from the sources below. Open them to check our work.

1
Primary source · Accessed September 30, 2026TileLang Ascend 950 Backend TileLang on GitHub
2
Research · September 30, 2026DeepSeek partners with Huawei to develop chip programming tools, reducing reliance on Nvidia Reuters
3
Research · September 30, 2026China’s DeepSeek open-sources tools to help Huawei chips supplant Nvidia in AI South China Morning Post

How we checked this story

We read TileLang’s own project pages and compared Reuters and the South China Morning Post, which both cite DeepSeek’s WeChat post. We could not read the post itself. The speed charts are the project’s own, so we make no claim about performance.

Story mapSOFTWARE STACK

How DeepSeek Tools for Ascend connects

TileLang
DeepSeek Tools for Ascend DeepSeek engineers added Ascend support , connected to TileLang
Ascend 950
DeepSeek Tools for Ascend a supernode of 128 Ascend 950 chips , connected to Ascend 950
Atlas 960E SuperPoD
DeepSeek Tools for Ascend two weeks after Huawei’s new AI systems , connected to Atlas 960E SuperPoD
Also in 1 other story

Picked by shared companies, people, places, and topics.

Chips & InfrastructureSep 19

Huawei’s new AI system replaces thousands of optical modules

The numbers still need outside tests.

The Atlas 960E SuperPoD links up to 4,096 AI processors with near-packaged optics. Huawei claims lower power use and higher reliability, while the first Ascend 960 chips are planned for 2027.

5 min read 3 sources

Chips & InfrastructureSep 28

China may let firms buy Nvidia’s RTX Pro 5500, report says

The Information reports the ministry asked ByteDance and Alibaba about their purchase plans. Reuters could not verify the report, and the card is not on sale yet.

China’s tech ministry asked companies including ByteDance and Alibaba to report their plans to buy Nvidia’s RTX Pro 5500, The Information reported on September 27. It told some of them that it intends to approve the purchases, according to the report. Reuters could not verify the report. An Nvidia spokesperson said US firms remain restricted by outdated US export controls and by China’s own limits on imports.

4 min read 4 sources

Chips & InfrastructureSep 23

Alibaba calls its Zhenwu V900 China’s most powerful AI chip

It has published no tests to prove it.

At its Apsara conference on September 22, Alibaba showed the Zhenwu V900, a chip for training and running AI. It says the chip has three times the performance of May’s M890 and will be mass-produced from early 2027. It also plans Qwen models with 5 to 10 trillion parameters.

5 min read 7 sources

Latest news

All news

Policy & LawOct 1

California bars firing or disciplining workers by AI alone

Newsom signed the No Robo Bosses Act with a dozen other bills on September 30. From July 2027, a human must check any discipline or firing that leans mainly on an automated system, and workers must be told in writing.

Governor Gavin Newsom signed SB 947, the “No Robo Bosses Act”, on September 30, along with 12 other bills his office grouped as AI and worker protections. From July 1, 2027, California employers may not rely solely on an automated decision system to discipline or fire workers. When such a system was the main basis, a human must corroborate the decision, and the worker must get a written notice.

4 min read

Safety & SecurityOct 1

OpenAI blames Moonshot-linked users for a copying campaign

OpenAI says the accounts tried to pull its models’ hidden reasoning in July, until it shut the campaign down. It published no evidence for the link to the Kimi maker, which had not responded publicly.

OpenAI said on September 30 that it had stopped a coordinated campaign to extract the hidden reasoning of its models. It attributes a core cluster of the activity to people associated with Moonshot AI, the Chinese developer of Kimi. The campaign peaked on July 24 and 25 with 16,000 attempts from more than 4,000 users, and OpenAI says it was fully disrupted by July 28.

3 min read

Safety & SecurityOct 1

DeepMind watermarks AI-designed proteins for biosecurity

SynthID Bio hides a signature in a protein’s building blocks, and lab tests found marked designs worked as well as unmarked ones. It could help DNA makers tell designs from trusted AI tools apart from the rest.

Google DeepMind introduced SynthID Bio on September 30, a way to watermark proteins designed by AI without breaking them. In lab tests on three targets, watermarked protein binders matched unwatermarked ones on how often they worked, how tightly they bound and how varied they were, DeepMind says. It is sharing the code, lab data and model weights with researchers.

4 min read

Policy & LawOct 1

Connecticut’s AI rules on layoffs and whistleblowers start

From October 1, employers must say whether AI played a part in mass layoffs, and makers of the biggest AI models may not punish staff who warn of catastrophic risks. AI subscriptions and AI-made media get new rules too.

Large parts of Connecticut’s new AI law took effect on October 1. Employers that file mass-layoff notices must now tell the state whether the layoffs relate to their use of AI. Developers of the biggest AI models may not retaliate against staff who report catastrophic risks, and AI subscription services must spell out their limits before customers pay.

4 min read

BusinessOct 1

ElevenLabs is now valued at $22 billion after a share sale

Employees sold $300 million of stock at twice February’s valuation, and the company raised no new money. ElevenLabs says businesses now bring in 55% of its revenue, and voice agents are its fastest-growing product.

ElevenLabs, the AI voice company, said on September 30 that a $300 million share sale by employees values it at $22 billion. Wellington and T. Rowe Price led the deal, and investors including EQT, Goldman Sachs and Singapore’s GIC bought in for the first time. The valuation is double the $11 billion of its funding round in February.

3 min read

What DeepSeek released

The role of TileLang

What is not known yet

Sources

Related stories

Latest news