Last mentioned: 1d ago
SK Hynix U.S. listing begins trading
South Korean memory giant SK Hynix debuts on the Nasdaq, seen as a crucial test of investor appetite for chip IPOs amid the AI rotation.
News report reveals the discussions
Anonymous sources leak details of the proposed restrictions, sparking speculation about China's AI containment strategy.
Chip stock selloff hits Nasdaq
PHLX chip index drops 4.65% as Samsung's record earnings fail to satisfy, Micron falls 4.7%, Sandisk loses 7.3%, and DeepSeek AI chip news adds competitive fears.
Report: DeepSeek developing own inference chip
Multiple sources report that DeepSeek is designing an in‑house AI inference chip to gain hardware independence from Nvidia and Huawei.
First Funding Round Closes
DeepSeek completes a $7.4 billion round at a valuation above $50 billion. Founder Liang Wenfeng contributes $3 billion, the government fund invests $150 million, and other investors commit under a 5-year lock-up through a limited partnership.
OpenAI unveils Jalapeno inference chip
OpenAI announces its first custom inference chip, developed with Broadcom, entering the custom silicon race.
Beijing holds meetings on AI access restrictions
Chinese Ministry of Commerce officials meet with Alibaba, ByteDance, and Z.ai to discuss limiting overseas access to advanced AI models, including criminalizing leaks and restricting foreign investment.
DeepSeek V4 optimized for Huawei Ascend chips
DeepSeek introduces its V4 AI model optimized for Huawei's Ascend hardware; Huawei confirms Ascend chips also used to train V4-Flash.
Expected Reveal
Anticipated official announcement and technical paper release from DeepSeek.
Market Speculation
Major tech outlets report on the 'mystery model' and its likely link to DeepSeek.
Beijing Approval
China's regulatory authorities grant approval for Nvidia to resume H200 sales; Jensen Huang confirms supply chain reactivation.
Supply Chain Activation
CEO Jensen Huang confirms purchase orders are being filled and the supply chain is 'firing up'.
Beijing Approval
Nvidia receives formal clearance from Beijing to resume H200 sales to many Chinese customers.
Performance Spike
Testers report top-tier results in reasoning and coding benchmarks.
Initial Sightings
Anonymous model placeholders appear on blind-testing platforms.
Expected Model Launch
DeepSeek is anticipated to release the model trained on Blackwell chips.
US Export License
US authorities grant a license for small amounts of H200 products to specific China-based entities.
Blackwell Usage Confirmed
U.S. official reveals DeepSeek trained its next model on restricted Blackwell hardware.
Anthropic Allegations
Anthropic accuses DeepSeek of harvesting data from Claude models via distillation attacks.
US Export License
The US government grants a license allowing limited exports of H200 products to specific China-based entities.
The Alibaba-led campaign represents a massive API abuse operation, deploying 25,000 fraudulent accounts to exfiltrate over 28.8 million Claude model responses, highlighting critical weaknesses in AI service security and the need for advanced threat intelligence sharing.
China's plan to restrict foreign investment in domestic AI startups threatens a funding ecosystem that has fueled the rise of companies like Z.ai and DeepSeek. VCs brace for a chill as government proposes national security vetting for new AI ventures.
SaaS developers who built services on top of China's cost-efficient AI models may need to repatriate to more expensive western APIs. A sudden restriction could raise cloud AI costs by up to 40% and disrupt product roadmaps.
Beijing weighs making AI model theft a national security offense and curbing foreign investments in AI startups, prompting urgent compliance reviews for global firms. Legal experts warn of broad extra-territorial reach and trade law implications.
China’s plan to restrict advanced AI model exports jeopardizes the open-source momentum that made DeepSeek R1 a global sensation. Developers fear a Balkanized AI future where each superpower hoards its best models.
China is considering restricting overseas access to its top AI models, which could upend the global AI market. For AI developers and enterprises that rely on inexpensive Chinese models like DeepSeek and Alibaba’s Tongyi Qianwen, this would mean sharply higher costs and potential supply chain disruption. The move underscores the escalating AI arms race and the weaponization of technology access.
Chinese AI startup DeepSeek is developing its own inference chip, a strategic shift that could lower costs for deploying its efficient models. The move intensifies competition in China's $50B AI chip market and signals a trend toward vertical integration among AI-first companies. Challenges remain, including manufacturing access and design expertise.
DeepSeek’s development of an inference chip marks a strategic shift with profound implications for semiconductor procurement and supply chain resilience. By reducing reliance on Nvidia and Huawei, the startup is quietly building a partner network spanning design, foundry, and memory — reshaping China’s $50B AI chip landscape.
DeepSeek's development of in-house AI inference chips signals a potential shift in semiconductor procurement, reducing reliance on Nvidia and Huawei while aligning with China's domestic chip push amid US export restrictions.
DeepSeek, China's AI champion, is reportedly developing its own inference chips—a high-stakes expansion that could redefine its valuation and challenge VC-backed AI hardware startups. The move signals a new phase in the AI chip race.
Chinese AI leader DeepSeek is developing its own inference chips, mirroring moves by OpenAI, to reduce Nvidia reliance. This shift underscores the growing importance of inference hardware in the AI ecosystem.
The Nasdaq tumbled 1.16% and the PHLX Semiconductor Index plunged 4.65% as Samsung’s earnings failed to impress overbought investors, triggering a rotation out of AI chip stocks. With Micron down 4.7% and Sandisk off 7.3%, the selloff raises fresh questions about stretched valuations. Upcoming SK Hynix listing will test whether global demand for AI hardware shares has peaked.
A brutal 4.65% drop in the PHLX chip index on July 7 was exacerbated by reports that Chinese startup DeepSeek is developing its own AI chip, posing a new competitive threat to Nvidia and Huawei. Even Samsung’s record earnings couldn't stop the rout, as the AI hardware boom faces a valuation reality check and rising global rivalry.
Meituan's LongCat-2.0 claims to be the first trillion-parameter model trained end-to-end on a 50,000-chip domestic compute cluster, matching Google's Gemini 3.1 pro. This achievement demonstrates that Chinese AI chips can now support large-scale training, intensifying the global AI race and loosening Nvidia's grip on advanced AI hardware.
Meituan's LongCat-2.0, the first trillion-parameter LLM trained entirely on a 50,000-chip domestic cluster, signals a breakthrough for Chinese cloud and SaaS providers. The achievement reduces dependence on sanctioned Nvidia chips and enables sovereign AI infrastructure, potentially accelerating AI-powered services across China's digital economy.
A Washington Post investigation exposes significant left-wing political bias in leading AI models, with OpenAI's GPT-5.5 showing an 80% left-leaning response rate. Google's Gemini demonstrates over 90% balanced answers, highlighting feasible neutrality in AI systems. This raises urgent questions for AI developers regarding training data, value alignment, and user trust.
Shanghai Stock Exchange relaxes IPO rules for unprofitable LLM developers, setting a 4 billion yuan anticipated market cap threshold and requiring 'clear commercialisation arrangements'. Legal advisors must scrutinize the new qualitative listing standards and their impact on pre-IPO structuring for Chinese AI labs.
The Chinese AI lab's first external fundraising round brings in $7.4 billion under a structure that keeps founder Liang Wenfeng in the driver's seat and imposes a five-year lock-up on investors. The deal sets new precedents for startup governance and the venture capital playbook in deep tech.
Nvidia CEO Jensen Huang has reframed AI intelligence as a tradeable commodity known as 'tokens,' likening data centers to modern factories. China is positioning itself to dominate this new 'tokenomics' landscape by leveraging its massive power infrastructure and a wave of ultra-low-cost models to drive global token exports.
Nvidia CEO Jensen Huang has defined the 'token' as the fundamental commodity of the AI era, transforming data centers into 'AI factories.' Meanwhile, China is positioning itself to dominate this new economy by leveraging its massive power infrastructure and low-cost model production to facilitate global 'token exports.'