DeepSeek Is Building Its Own AI Chip, Report Says
DeepSeek is developing its own AI inference chip to reduce reliance on Nvidia and Huawei, per a Reuters exclusive, as it reportedly raises $7B at a $52–59B valuation.
TL;DR — DeepSeek is developing its own AI inference chip to cut reliance on Nvidia and Huawei, according to a Reuters exclusive — as the Chinese AI startup was also reported to be raising $7 billion at a $52–59 billion valuation.
The Chinese AI lab that shook the industry now wants its own silicon. On July 7, 2026, Reuters reported DeepSeek is developing an in-house AI chip.
What was reported
Per a Reuters exclusive, DeepSeek is developing an AI chip optimized for inference — running models, not training them — to cut its reliance on Nvidia and Huawei silicon. The company has reportedly been in talks with chip-design, foundry and memory partners for about a year, hiring chip engineers quietly. Separately, DeepSeek was slated to raise $7 billion at a $52–59 billion valuation, reversing its earlier no-outside-investment stance.
| Detail | |
|---|---|
| Chip focus | Inference (not training) |
| Goal | Reduce Nvidia / Huawei reliance |
| Reported raise | $7B at $52–59B valuation |
| Sourcing | Reuters, three people familiar |
Important caveat: the report is anonymously sourced, and DeepSeek did not comment on the record.
What they said
"DeepSeek has almost no chance of selling silicon outside of China unless it gets access to leading edge manufacturing." — Richard Windsor, analyst, Radio Free Mobile
Why it matters
- Export controls bite. US rules bar China from Nvidia’s most advanced chips, pushing labs toward homegrown silicon.
- Everyone is verticalizing. OpenAI, Meta and Anthropic are all building custom inference chips too.
- Manufacturing is the ceiling. Without leading-edge fabrication, a Chinese-designed chip stays largely domestic.
FAQ
Is DeepSeek making its own AI chip?
According to a Reuters exclusive published July 7, 2026, DeepSeek is developing an AI chip optimized for inference to reduce reliance on Nvidia and Huawei. The report is based on three anonymous sources; DeepSeek did not comment on the record.
Why would DeepSeek build its own chip?
US export controls prevent China from buying Nvidia’s most advanced chips. Designing an in-house inference chip would reduce DeepSeek’s dependence on restricted foreign silicon — though without leading-edge manufacturing access, its reach would be mostly domestic.
Sources
- Reuters via U.S. News — China’s DeepSeek developing its own AI chip
- Bloomberg — DeepSeek developing own AI chip, Reuters says
Image: “DeepSeek logo” by DeepSeek, MIT License, via Wikimedia Commons.
← Back to all posts