Yazılar

Huawei unveils chip and computing power roadmap in challenge to Nvidia

/in /tarafından

Huawei on Thursday publicly detailed its long-term semiconductor ambitions for the first time, pledging annual upgrades to its Ascend AI chips and unveiling plans for powerful computing systems designed to rival Nvidia (NVDA.O).

At its annual Huawei Connect conference in Shanghai, rotating chairman Eric Xu said the company will follow a one-year release cycle that doubles computing power with each generation. Huawei also revealed it has developed its own high-bandwidth memory, a technology currently dominated by South Korea’s SK Hynix (000660.KS) and Samsung Electronics (005930.KS).

Xu said Huawei’s upcoming Atlas 950 supernode, slated for launch in late 2026, will connect 8,192 Ascend chips, while the Atlas 960 in 2027 will link 15,488 chips. He claimed these systems will “far exceed” competitors on key performance metrics. Huawei will also release new Kunpeng server chips in 2026 and 2028.

The roadmap underscores China’s push to reduce reliance on U.S. suppliers amid intensifying trade and technology tensions. This week, Beijing accused Nvidia of antitrust violations and ordered major Chinese firms to halt purchases of its AI chips, according to the Financial Times. The moves come just before a scheduled meeting between Presidents Donald Trump and Xi Jinping.

Huawei first entered chipmaking in 2018 but retreated from public disclosures after U.S. sanctions in 2019 restricted its access to advanced chipmaking tools. Since then, analysts say the company has become a leader in China’s domestic semiconductor push. Its current AI flagship, the Ascend 910C, launched earlier this year, with the Ascend 950 due in 2025, followed by the 960 in 2027 and the 970 in 2028.

Despite these advances, engineers at Chinese tech firms acknowledge Nvidia’s chips remain more powerful. U.S. export controls continue to limit Huawei’s access to cutting-edge manufacturing technology, though Washington recently eased some restrictions on downgraded Nvidia chip sales.

“Huawei is leveraging its networking strengths and China’s power supply advantages to push aggressively into supernodes, offsetting lagging chip manufacturing,” said Wang Shen of Omdia.

Chinese semiconductor stocks rose 3.4% on Thursday after reports of the Nvidia purchase ban. Beijing’s foreign ministry responded cautiously, saying China remains open to dialogue to stabilize global supply chains.

Nvidia spends $900M to hire Enfabrica CEO and license startup’s tech

/in /tarafından

Nvidia (NVDA.O) has spent more than $900 million to bring on Enfabrica CEO Rochan Sankar, along with other employees from the AI hardware startup, while also licensing the company’s technology, CNBC reported Thursday.

The deal, paid in cash and stock, closed last week, and Sankar has already joined Nvidia, according to people familiar with the arrangement.

Enfabrica, based in Silicon Valley, is addressing a major bottleneck in artificial intelligence: how to connect tens of thousands of chips into a network that can operate as a single, unified computer. Without efficient networking, even high-performance chips like Nvidia’s can sit idle while waiting for data. The startup’s technology reportedly enables up to 100,000 AI chips to be linked before network slowdowns occur.

Nvidia declined to comment on the report, and Enfabrica did not immediately respond to inquiries. Founded by former Broadcom (AVGO.O) and Alphabet (GOOGL.O) veterans, Enfabrica has raised $260 million in venture capital and in July launched a chip-and-software system designed to reduce memory costs in data centers.

The move mirrors recent strategies by Big Tech rivals. In June, Meta (META.O) acquired a 49% stake in Scale AI and brought CEO Alexandr Wang into its AI leadership, while Google hired key staff from Windsurf, an AI code generation startup courted by OpenAI.

China investors stay bullish on Cambricon despite index reshuffle

/in /tarafından

Cambricon Technologies, often dubbed China’s Nvidia, faces more than 8 billion yuan ($1.1 billion) in passive outflows due to a quarterly rebalancing of the STAR50 Index, but analysts say investor confidence in the AI chipmaker remains intact.

The company’s stock, which more than doubled in August, exceeded the 10% cap for individual weightings in the tech-heavy index. Though Cambricon shares fell 14% last week on profit-taking and rebalancing fears, they have since rebounded 10%, hovering near record highs.

Valuations are eye-watering—Cambricon trades at 521 times earnings, compared with Nvidia’s multiple of 50—but Beijing’s push for tech self-sufficiency, the DeepSeek AI breakthrough, and large-scale investments by Alibaba, Tencent, and Baidu continue to fuel the rally.

“Maybe some investors will use it as a reason to take profit, but I don’t think that will affect the long-term trend,” said Shihao Li, analyst at CLSA. Gavekal’s Tilly Zhang added that optimism is growing that China’s AI sector has entered a “self-sustaining cycle of rising investment and higher profitability.”

Cambricon’s fundamentals have helped power the surge. First-half revenue jumped to 2.9 billion yuan ($407 million) from just 64.8 million yuan a year earlier, swinging to a 1 billion yuan profit. The company forecasts 5–7 billion yuan in operating revenue for 2025.

Still, risks remain. Some fund managers warn of a speculative bubble, while others argue that growth potential tied to China’s strategic need to replace foreign AI chips may justify lofty valuations.

Broader Chinese markets are riding the same wave. The CSI AI Index is up 60% this year, far outpacing the 15% gain in the CSI300, and the Shanghai Composite has hit levels not seen in a decade.

The spotlight now shifts to whether Cambricon can sustain profitability and meet surging demand for AI chips—critical to maintaining its role as the flagship of China’s AI boom.