China's AI Chip Push Gives Xi Leverage in Washington
New processors from Huawei and Alibaba landed days before Xi Jinping's White House visit, but analysts say China still depends on U.S. chips.
Technology Editor

China's AI chip blitz arms Xi Jinping with a direct message for U.S. export controls
WASHINGTON: Chinese President Xi Jinping arrived for his first U.S. state visit in more than a decade backed by a run of new chip and AI model launches from Chinese companies, according to a Sept. 24 report by CNBC. The releases strengthen Beijing's argument that U.S. export controls can slow its AI industry but not stop it.
Xi's visit to the White House is his first in 11 years and follows President Donald Trump's state visit to Beijing in May. The summit got under way in Washington on Thursday, Sept. 24.
Huawei and Alibaba Time Their Launches
The timing looked planned. At its annual Huawei Connect conference last week, Huawei said its Ascend 960DT chip will ship in the first quarter of 2027, three quarters ahead of schedule. TradeFlock covered that accelerated Ascend launch when it was announced. On Tuesday, at the Apsara Conference in Hangzhou, Alibaba's chip unit unveiled the Zhenwu V900, which it says triples its predecessor's performance and goes on sale in early 2027.
DeepSeek, the startup whose low-cost model rattled markets last year, has reportedly lined up the largest known cluster of Huawei chips built to date.
"For Xi, the leverage is not 'we have caught up with you.' It's 'you can't simply technologically choke us off,'" said Lizzi Lee, a fellow at the Asia Society Policy Institute. George Chen of The Asia Group called the timing a "deliberate" confidence boost. "You need to have cards when you sit at a table, and Xi has many cards," he said.
Clusters Make Up for Weaker Chips
Chinese firms are trying to offset weaker individual chips by linking thousands of them over faster connections. Sigrid Wang of Hutong Research said there "does seem to be genuine acceleration" in Huawei's production timeline.
The approach has limits. Huawei's latest SuperPoD architecture links far fewer processors than planned, and each chip delivers roughly half the computing power of Nvidia's. Huawei Chairman Eric Xu has acknowledged the company may not make enough chips to meet domestic demand.
The Revenue Gap With U.S. Labs
Skeptics say the fundamentals have not moved. "Chinese chip companies haven't made any groundbreaking progress in the past few years," said Xiaomeng Lu of Eurasia Group.
Chris Miller, the Tufts University professor who wrote "Chip War," said China often announces projects around high-level meetings to show it is "catching up." U.S. models still lead on benchmarks and profits, he said, and American firms such as Anthropic and OpenAI make 50 to 100 times the revenue of comparable Chinese companies. A separate estimate by Rhodium Group put OpenAI and Anthropic at ten times the revenue of all Chinese AI models combined.
Miller added that Chinese firms remain short of the chips they need and rely on smuggling or on data centers outside China. "China is still highly dependent on the U.S. for advanced chips," he said.
Export Controls Stay Off the Summit Agenda
Neither side seems ready to negotiate on chips. Treasury Secretary Scott Bessent and U.S. Trade Representative Jamieson Greer met Vice Premier He Lifeng in New York before the summit and set up a formal AI channel for incident warnings, a step Beijing later confirmed as the two countries' first AI talks. Greer said export controls on advanced chips and chipmaking equipment were not on the agenda.
Beijing's priorities may lie elsewhere. Lyle Morris of the Asia Society Policy Institute said Xi is likely to press Trump on Taiwan and could offer help on Iran in return, including restoring free movement of shipping through the Strait of Hormuz.
For chip suppliers and their customers, that leaves the current rules in place. "The recalibration of U.S.-China tech relations takes a lot of substantial work," Lee said. "Not a summit deliverable, I am afraid, at least not this time."
James Whitaker
Technology Editor
Reports on semiconductors, cloud infrastructure, and the industrial politics of AI.







