Huawei's AI chips lag behind Nvidia due to stability issues

Huawei is sending engineers to help its customers to resolve THIS problem
An undated image of AI chip. — Unsplash
An undated image of AI chip. — Unsplash

Huawei artificial intelligence (AI) chips are purportedly facing stability issues and this is likely to be a piece of good news for Nvidia. China is finding ways to reduce dependence on United States (US) suppliers.

However, these efforts are ruined by flawed software in Huawei’s chip core. According to the Financial Times, Huawei customers complained about the buggy performance of Ascend AI chips, resulting in stability issues. An AI engineer said that the processor continuously crashes leading to difficult and unstable work.

A significant reason behind the stability issues of Huawei AI chips is its buggy software recognised as “Cann”. The Huwei-developed software was released as an alternative to Nvidia’s Cuda. It was reported to help in improving the chip performance for developers.

However, it looks like the software has run into a strange bug that impacted the AI chip to perform poorly in different training models. Moreover, it displayed a slower inter-chip connection.

Read more: Huawei to counter iPhone 16 series launch with its triple-foldable smartphone

Several Huawei employees also poured out their opinions and said: “When random errors occur, it is very difficult to find out where it comes from due to poor documentation. You need talented developers to read the source code to see what the issue is, which slows everything down. The coding is imperfect.”

Despite all these issues, Huawei has begun sending engineers that can help its customers to handle the problem. The report disclosed that experienced developers transfer the training code written on Cuda into Cann to ease the overall operation.

Some major companies, such as Baidu, Tencent, and iFlytek now have a new team of engineers to solve the AI chip-related problems. A tech analyst spotlighting this issue said: “Huawei’s advantage over Nvidia is it can work closely with its customers. Unlike Nvidia, it has a large team of engineers to help solve clients’ problems and get them to transition to their hardware.”

Apart from sending teams, the company also opened an online portal that can help customers with constant services and feedback to improve its buggy software.