Chinese AI startup DeepSeek is developing its own AI chip to power AI systems, according to a report published on July 7, 2026 . Citing three people familiar with the matter, the chip is aimed at inference and intended to reduce reliance on chips from Nvidia and Huawei.
July 7, 2026 · DeepSeek
DeepSeek Is Building Its Own AI Chip
The Chinese AI startup is quietly developing an inference-focused chip — work reportedly begun a year ago — in a bid to cut its reliance on both Nvidia and Huawei silicon.
~1 yr
Time the chip effort has quietly been under way
Inference
The target — not model training
~$50B
Size of China's AI chip market DeepSeek is entering
China's AI Chip Market — Share of ~$50B
~50%
Others — Alibaba, Baidu & more
Huawei dominates roughly half the market; Alibaba and Baidu already build their own chips — and DeepSeek's entry could further complicate Huawei's position.
The Hardware Journey
Nvidia H800
V3 / R1 trained under $6M
→
Huawei Ascend
V4 optimized, Apr 2026
→
Own Chip
Inference · early stage
Still Unannounced
No product name, specifications, or release timeline has been disclosed — and the plan could change as work progresses.
In talks with chip design, foundry and memory firms; hiring chip engineers
Backdrop: U.S. export controls and Beijing's push for domestic silicon
As the report circulated, Nvidia shares fell more than 2% in premarket trading
Strengths
ChatGPT-class performance at a fraction of the cost, with strengths in math and coding and growing adoption on domestic clouds.
Concerns
U.S. officials cite alleged export-control evasion and data-sharing; the Huawei shift surfaced stability and software-maturity issues.
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article…