Skip to content
Back to mrpetzai.de Tuesday, November 3, 2026 Edition 13 · 7 stories
Mr. Petz AI MrPetzAI AI News The weekly briefing for AI decision-makers
CN China

Huawei's chip roadmap narrows little of Nvidia's lead by 2030.

Huawei will stay about four years behind Nvidia on AI chip performance through 2030, Epoch AI finds.

Infrastructure Executive
AI-GENERATED
Text size

Epoch AI compared Huawei's Ascend roadmap with Nvidia's chip development. Huawei will produce roughly 25 times less total AI compute than Nvidia in 2026, and US export controls on manufacturing and memory bandwidth are unlikely to close that gap by 2030.

Epoch AI compares Huawei's Ascend 950 chip with Nvidia's current generation by looking at total compute, meaning the number of chips produced multiplied by each chip's performance. The Ascend 950 delivers roughly seven times less compute throughput than Nvidia's B300 and only half the performance of the 2022-era H100. On volume, Epoch AI expects Huawei to produce about 1.5 million Ascend chips in 2026, compared with roughly 6 million chips from Nvidia. Combined, that gives Huawei about 25 times less total compute than Nvidia.

Huawei's manufacturing partner SMIC cannot pack transistors as densely as Nvidia's partner TSMC, because it lacks access to advanced lithography equipment. Huawei plans to close part of that gap by 2030 through bigger chips, an architecture optimized for lower-precision math, and a vertical stacking technique called LogicFolding. Epoch AI estimates that despite substantial progress, Huawei will still trail Nvidia by roughly four years in both chip performance and production volume by 2030. For companies relying on Chinese supply chains, that points to a lasting compute gap that should factor into procurement and diversification decisions.

It remains unclear how the picture changes if China matures domestic lithography tools sooner than expected, or if US export controls ease. Nvidia's own progress is not guaranteed either, for instance if its Feynman generation, expected in 2028, slips. For companies dependent on Chinese AI hardware, the open question is whether a four-year performance gap even matters for their workloads, or whether lower raw performance is offset by price and availability.

What this means for decision-makers

  • Check whether available Chinese chip compute meets your workload needs despite a four-year gap to Nvidia.
  • Monitor Chinese lithography progress, since it could close the performance gap faster than currently expected.
  • Hedge Nvidia supply chain risk through alternative sourcing without assuming Huawei will catch up soon.
Source Epoch AI

This story was produced automatically from the source named above and checked by software before publication. The image is symbolic and shows neither the event nor a real person. How this paper is made

Free subscription

Get the whole edition by email. Free of charge.

Every Tuesday morning, seven documented stories with what each one means for your decisions. One click to unsubscribe.

Subscribe free of charge

No costs, no advertising, no forwarding of addresses.

Free subscription

The whole edition free of charge by email.

Seven documented stories from research, public authorities and standardisation – and what they mean for executives, marketing, HR and sales. No costs, one click to unsubscribe.

Subscribe free

Double opt-in: nothing is sent before you confirm the link in the email.

Editorial principles

Always the original source

Every story names its source and links to it directly. We do not pass on information we cannot trace.

Interpretation, not excitement

Every story states what it means for decisions in your company – concretely, not as a buzzword.

Organised by country

The United States sets the pace, Germany sets the frame. The other markets follow by actual AI activity.