Unlock the Editor’s Digest totally free
Roula Khalaf, Editor of the FT, selects her favorite tales on this weekly e-newsletter.
Huawei’s founder stated the US was exaggerating the Chinese language chipmaker’s capabilities and performed down his firm’s know-how amid commerce talks between Beijing and Washington that embody discussions of export controls.
In a uncommon interview with China’s state-run Individuals’s Each day on Tuesday, Ren Zhengfei stated Huawei’s Ascend chip, the primary rival to Nvidia’s merchandise in China, “still lags behind the US by one generation”. He added that the “US has exaggerated Huawei’s capabilities — we’re not that strong yet”.
Ren’s feedback come as Nvidia chief government Jensen Huang has sounded the alarm in latest weeks on Huawei’s progress with its synthetic intelligence chips, saying Washington’s limits on the US chipmaker promoting to China have given rise to a “formidable” competitor that threatens to undermine American dominance of AI know-how.
The US and China started a brand new spherical of commerce talks on Monday in London, which included dialogue of Washington’s export controls on key know-how.
Throughout the first spherical of talks in Geneva, the US didn’t talk about export controls. Nonetheless, Beijing’s latest use of restrictions on some crucial uncommon earths and minerals utilized in automotive manufacturing — which threatens to close down manufacturing unit traces within the US, Europe and Japan — has introduced the difficulty into commerce discussions.
Huawei has benefited from Washington’s ban on shipments of Nvidia chips to China as Chinese language tech giants have accelerated purchases of Ascend chips and made preparations to undertake Huawei’s know-how.
Nonetheless, nearly all of Chinese language AI corporations, together with DeepSeek, use Nvidia chips to coach the big language fashions that energy AI instruments. Home options are more and more used for the much less complicated job of calling upon the fashions to generate responses in instruments similar to chatbots.
Analysts and Huawei researchers have beforehand complained about technical glitches in utilizing the corporate’s chips for coaching LLMs, citing problem in making the chips work collectively and distributing computing workloads throughout them.
Ren on Tuesday hinted the corporate had made strides to resolve these points, saying Huawei may “compensate” for poorer efficiency by means of cluster computing, which entails linking a number of chips to spice up AI server energy.
“Using clustering and stacking, our computing results are comparable to the world’s best,” he stated.
Huawei’s new CloudMatrix 384 AI server is the centrepiece of its technique to compete with Nvidia by assembling numerous chips to bolster bandwidth capability and deal with extra information. The server connects 384 AI processors utilizing Huawei’s optical know-how.
A number of Huawei prospects are already testing the CloudMatrix server, working with the corporate’s engineers to resolve points associated to the warmth generated by the big variety of chips working in unison and the machine’s weight, stated an individual concerned within the testing.
Huawei’s greatest problem was creating “an ecosystem” of builders that may use its platform to construct AI fashions, the individual added. Considered one of Nvidia’s benefits is its software program platform Cuda, which builders say is simple to make use of.
Ren stated Huawei was investing Rmb180bn ($25bn) a 12 months on analysis and improvement, with Rmb60bn going into primary analysis not geared toward creating merchandise however making breakthrough discoveries.
Ren stated China had distinct benefits when constructing out its technological capabilities.
“AI depends on abundant electricity and advanced network infrastructure,” he stated. “China’s power generation and grid systems are world-class. Our telecoms infrastructure is the most advanced in the world.”
Further reporting by Demetri Sevastopulo in Washington
