China's Chip Industry Achieves Breakthroughs: CXMT's IPO Catapults It into Global Top Three, Huawei's AI Chips Now Capable of Training Large Models
Nashnova编辑部
CXMT's stock has surged 466% since its July IPO, making it China's most valuable A-share company, while Huawei's AI chips can now train complex large language models — but history shows lasting breakthroughs require both policy support and a powerful anchor customer.
How did a memory-chip maker become China's most valuable listed company?
CXMT's stock has risen 466% since its July IPO, overtaking major oil companies and state banks to become China's most valuable A-share company.
The IPO ranked third globally this year, behind only SpaceX and SK Hynix.
This means → the market is pricing CXMT not as a chipmaker but as a national strategic asset — a bet on China's ability to replace imported memory chips at scale.
Where do Huawei's AI chips stand now?
Huawei's AI chips can now train complex large language models and support autonomous driving and humanoid-robot motion control.
Chinese firms are advancing on two parallel tracks: developing some chipmaking equipment domestically + retrofitting older foreign-made tools to boost performance.
In plain terms = U.S. restrictions on the latest equipment exports have pushed China into a dual track — upgrade what you have, build what you can't buy.
What does history say breakthroughs require?
Bloomberg identifies a historical pattern: semiconductor breakthroughs need two elements at once — effective state industrial policy and at least one powerful strategic customer.
Texas Instruments rose on Cold War federal contracts + IBM mainframe partnerships. Intel locked in its position as IBM PC's sole chip supplier + federal subsidies. TSMC was built on decades of Taiwanese industrial policy + deep ties with Apple and Nvidia.
This reflects a pattern tested repeatedly: government money alone is not enough, and customer demand alone is not enough — remove either element and the breakthrough stalls.
Are both conditions coming together in China?
Policy side: Beijing has made breaking specific technology bottlenecks a top semiconductor priority, with special focus on the links critical to China's AI strategy.
Customer side: Huawei and Tencent are playing the role IBM once did — acting as strategic demand anchors that pull domestic chipmakers toward higher capability.
This means → the two historical prerequisites are forming in parallel in China, and that is the most fundamental difference between this wave and the earlier spend-first approach.
Why did "Made in China 2025" fall short?
The "Made in China 2025" plan pledged roughly $150 billion to nurture domestic chips, but five years in, results were limited and multiple projects stalled.
The clearest example: Tsinghua Unigroup — once seen as China's national chip champion — ended up in bankruptcy restructuring.
In plain terms = the last round proved that policy funding without a real technology customer to pull demand forward does not grow an industry. Whether today's breakthroughs can last still depends on how fast genuine technical bottlenecks fall and how the international competitive landscape shifts.
Content is for reference only, not financial advice.