Chinese GPU Makers Debut on Shanghai Amid US Chip Curbs

chinese gpu ipo shanghai​

Moore Threads kicked off trading on the Shanghai Stock Exchange’s Star Market this Friday, securing its spot as China’s second-largest onshore IPO of 2025 by raising nearly 8 billion yuan, equivalent to about $1.13 billion. This debut valued the Beijing-based GPU designer at roughly $7.6 billion, reflecting massive investor confidence in China’s push for semiconductor self-reliance. Shares launched at 114.28 yuan each—the highest issue price for any A-share IPO this year—amid extraordinary retail demand that oversubscribed the offering by 4,000 times, resulting in a razor-thin final allotment rate of just 0.036 percent. This frenzy underscores Beijing’s aggressive acceleration of domestic alternatives to Nvidia, especially as ongoing US export restrictions continue to limit access to cutting-edge American technology.​

At the same time, MetaX Integrated Circuits opened subscriptions on the same day for its own Star Market listing, setting shares at 104.66 yuan and targeting 3.9 billion yuan in fresh capital. Both companies operate as “fabless” chipmakers, meaning they focus on designing high-performance GPUs while outsourcing the actual manufacturing to foundries. They position themselves squarely as viable homegrown rivals to Nvidia’s market-leading GPUs, stepping into a void created by US policies that have made prospects for American chip giants in China highly uncertain. Chinese regulators fast-tracked these listings—approving Moore Threads’ registration in a record 88 days and MetaX in 116 days—signaling strong governmental backing for strategic technologies vital to national security and AI ambitions.​

Market Opportunity Amid Export Controls

The timing of these debuts aligns perfectly with escalating US-China tech tensions, where export controls have effectively barred Nvidia’s most advanced AI chips from the Chinese market. Nvidia CEO Jensen Huang recently acknowledged in public statements that his company’s share of China’s advanced AI chip market has plunged from a dominant 95 percent to nearly zero, forcing it to pivot toward compliant but less powerful alternatives like the H20. According to detailed analysis from Bernstein Research, Nvidia still commanded 66 percent of China’s overall accelerator market in 2024, with Advanced Micro Devices (AMD) at 5 percent, Huawei’s HiSilicon unit at 23 percent, and emerging player MetaX scraping by with about 1 percent. These shifts highlight how US restrictions are reshaping the landscape, propelling local firms forward as Beijing invests heavily in its domestic supply chain. Projections indicate Chinese AI chip production could surge from 17 percent of global output in 2023 to as much as 55 percent by 2027, driven by policies favoring innovation in restricted sectors.​

MetaX Chairman Chen Weiliang outlined ambitious plans during a Thursday online presentation, touting the company’s flagship C600 GPU as delivering performance slotted between Nvidia’s widely used A100 and the more advanced H100 models. Full-scale production of the C600 is on track for the first half of 2026, with the next-generation C700 GPU—crafted entirely using China’s domestic supply chain—expected to go head-to-head with the H100 by the second half of that year. Moore Threads, frequently dubbed “China’s Nvidia,” complements this ecosystem with its own lineup, including the MTT S4000 tailored for AI training and inference workloads, as well as graphics-focused chips that target data centers, supercomputing, and even consumer gaming. These developments come as Chinese firms ramp up efforts to fill the gaps left by sanctioned imports, with state-backed initiatives pouring resources into R&D for everything from chip architecture to advanced packaging.​

Financial Performance and Outlook

Moore Threads delivered impressive top-line growth, reporting 780 million yuan in revenue across the first three quarters of 2025—a staggering 182 percent increase from the prior year—fueled by rising demand for its domestic GPUs in AI servers and high-performance computing. Despite this momentum, the company recorded a net loss of 724 million yuan over the same period, though it narrowed that deficit by 19 percent compared to 2024, thanks to scaling operations and cost controls. Founded in 2020 by James Zhang Jianzhong, a former Nvidia vice president who oversaw the US giant’s China operations for 14 years, Moore Threads brings deep industry expertise to its mission. Zhang’s background gives the firm credibility in navigating complex GPU design challenges, and management now projects achieving consolidated profitability as early as 2027. Proceeds from the IPO will primarily fund development of next-generation AI and graphics processors, alongside bolstering working capital to support rapid expansion in a competitive market.​

MetaX’s financial trajectory shows similar patterns of high growth amid heavy investments, with 743 million yuan in revenue for full-year 2024 but a substantial net loss of 1.4 billion yuan, reflecting the intense R&D spend required to catch up in AI chip tech. By the first half of 2025, revenue had skyrocketed 404 percent to 915 million yuan, while losses began to moderate as production ramped up—evidence of improving economies of scale. Led by executives with pedigrees from AMD and other global players, MetaX secured its IPO nod swiftly, planning to channel the 3.9 billion yuan raise into advancing high-performance GPU lines like the C600 and C700 series. Earlier products such as the C500 faced setbacks from US sanctions, but the company persists, betting on China’s burgeoning domestic ecosystem. Both Moore Threads and MetaX exemplify the broader trend: young Chinese chip startups burning cash on innovation today to dominate tomorrow’s market, backed by a government prioritizing tech sovereignty over short-term profits.​


Subscribe to Our Newsletter

Related Articles

Top Trending

Infographic illustrating the Retrieval-Augmented Generation (RAG) process for AI, showing data retrieval, context augmentation, and generation for a grounded response.
RAG Explained: How AI Systems Ground Their Answers in Real Data
FunnyMike Net Worth
FunnyMike Net Worth: Earnings From Instagram And TikTok
Digital security dashboard displaying key cybersecurity statistics for businesses alongside a glowing shield icon and network servers.
11 Cybersecurity Statistics That Should Worry Every Business
Mobile app for your business planning shown with app wireframes, performance analytics, budgeting notes, and multiple devices used to evaluate development decisions.
10 Questions to Ask Before Deciding on a Mobile App for Your Business
How AI Models Are Trained on Data And Why Data Quality Matters
How AI Models Are Trained on Data And Why Data Quality Decides Your Output

Technology & AI

Infographic illustrating the Retrieval-Augmented Generation (RAG) process for AI, showing data retrieval, context augmentation, and generation for a grounded response.
RAG Explained: How AI Systems Ground Their Answers in Real Data
Digital security dashboard displaying key cybersecurity statistics for businesses alongside a glowing shield icon and network servers.
11 Cybersecurity Statistics That Should Worry Every Business
Mobile app for your business planning shown with app wireframes, performance analytics, budgeting notes, and multiple devices used to evaluate development decisions.
10 Questions to Ask Before Deciding on a Mobile App for Your Business
How AI Models Are Trained on Data And Why Data Quality Matters
How AI Models Are Trained on Data And Why Data Quality Decides Your Output
AI tool feature bloat shown through a crowded AI workspace filled with extra tools, illustrating how added features can complicate a simple interface.
AI Tool Feature Bloat: Why AI Apps Keep Adding Things You Don’t Need

GAMING

Online Color Game Philippines
Online Color Game Philippines: What Every Beginner Should Know Before Playing
Ways to Reduce Game Development Costs
12 Ways Studios Cut Game Development Costs
NFT game development cost
How Much Does NFT Game Development Cost? A Realistic Budget Breakdown
Reasons Why You No Longer Need the Best Roblox AI Scripter
Forget Best Roblox AI Scripter: 10 Reasons Why You No Longer Need It
Blockchain Platforms for Game Development
The 9 Best Blockchain Platforms for Game Development

Business & Marketing

Mobile app for your business planning shown with app wireframes, performance analytics, budgeting notes, and multiple devices used to evaluate development decisions.
10 Questions to Ask Before Deciding on a Mobile App for Your Business
CAC Payback
Why CAC Payback Matters Far More Than Cheap Customer Acquisition
LTV to CAC ratio
The LTV to CAC Ratio: What It Is and Why Everyone Quotes It
effective meeting management
Top 10 Ways to Master Effective Meeting Management and Save Time
API cost management
The Business of APIs: Why Cost Management Decides a Tool’s Survival

EdTech & E-Learning

How Long Does It Take a Child to Learn the Alphabet
How Long Does It Take a Child to Learn the Alphabet? A Real Timeline
Games to Encourage Early Language Skills
I Tried 8 Games to Encourage Early Language Skills [One Flopped]
Active recall and spaced repetition
How to Study With Active Recall and Spaced Repetition: A Practical Guide
early math myths
8 Early Math Myths That Hold Kids Back
Bedtime Math
Bedtime Math: 7 Clever Ways to Boost Math Confidence

Software & Apps

Mobile app for your business planning shown with app wireframes, performance analytics, budgeting notes, and multiple devices used to evaluate development decisions.
10 Questions to Ask Before Deciding on a Mobile App for Your Business
Best Scheduling Tools
10 Best Scheduling Tools to Kill the Back-and-Forth
The Rebirth of ImagineLab Art
The Rebirth of ImagineLab Art: Inside the Unified AI Creative Platform in One Seamless Workspace
best minimalist apps
10 Best Minimalist Apps That Do One Thing Well
Best Voice-to-Text Apps
10 Best Voice-to-Text Apps for Capturing Ideas for Writers and Creators