Bill Gross warns on gold momentum as regional bank stocks tumble
Investing.com -- Huawei Technologies Co. has publicly revealed a three-year strategy to challenge Nvidia Corp.’s dominance in the artificial intelligence chip market, despite acknowledging its silicon cannot match Nvidia’s raw power and speed, according to a recent report from Bloomberg News.
During Thursday’s annual Huawei Connect conference, Rotating Chairman Eric Xu detailed the company’s technological vision in an unusually public display for the typically secretive Chinese tech giant. The presentation came just one day before US President Donald Trump and Chinese leader Xi Jinping held their second phone conversation in four months.
The Shenzhen-based company’s approach relies on networking capabilities, scale, and policy support to compensate for performance gaps. Huawei introduced its next generation of Ascend-branded AI chips alongside upgraded "SuperPod" designs - a term borrowed from Nvidia that refers to comprehensive data center platforms combining computing, storage, networking, software and infrastructure management technologies.
Huawei’s strategy involves connecting up to 15,488 of its AI chips using its newly unveiled UnifiedBus interconnect protocol. The company claims this technology enables data transmission between individual chips up to 62 times faster than Nvidia’s upcoming NVLink144 technology. For comparison, Nvidia’s current NVLink72 technology connects 72 Blackwell graphic processing units and 36 Grace central processing units.
This public roadmap represents a significant departure for Huawei, which has operated more discreetly since losing access to Taiwan Semiconductor Manufacturing Co. in 2020 due to US restrictions.
This article was generated with the support of AI and reviewed by an editor. For more information see our T&C.