News: Zhipu is evaluating its self-developed custom AI chip
Related
Anthropic has launched its self-developed AI chip and is in talks with Samsung regarding a customization plan.
According to Beating's monitoring, Anthropic is in talks with Samsung regarding a custom AI chip and has already begun early development work on its self-developed AI chip. Anthropic has also held talks with several chip design companies. --------------------------------- Click the original link below to join Beating's AI news channel on Lark, and receive 24/7 monitoring of global AI hot topics and news.
Zhipu plans to develop its own AI chip and has already contacted chip companies to accelerate efforts to address its computing power shortcomings.
According to sources familiar with the matter, Zhipu is evaluating the possibility of designing its own AI chip to address the rapidly growing demand for its GLM series of open-source large models and the shortage of computing resources caused by US export restrictions. Odaily has recently made preliminary contact with some Chinese chip design companies to explore cooperation in developing a customized AI processor optimized for its model operation scenarios. Zhipu is one of China's leading AI large-scale model companies, with its GLM series models attracting attention due to its open-source strategy and performance. As the application scale of large-scale models expands, high-performance computing resources have become a key bottleneck for AI companies. Developing its own chips is seen as a way to reduce dependence on external GPUs and improve computing efficiency. Currently, Zhipu AI has not yet made a final decision on whether to proceed with the chip project; the relevant plans are still in the early evaluation stage. (The Information)
Zhipu surged over 10% after reports indicated it was considering developing its own AI chips.
PANews reported on July 8 that, according to Jinshi News, Zhipu (02513.HK) rose by more than 10%, with a turnover of over HK$6 billion. The report stated that Zhipu is considering developing its own AI chip to cope with the rapid growth in demand for GLM models.
Last night's breaking news roundup! Fill in the information gaps in 1 minute | 7/8
1. The US military launched a powerful strike against Iran, 4-5 times larger than 10 days ago, escalating tensions in the Middle East. 2. The US lifted oil sanctions on Iran, causing Brent crude to surge 5.68% to $75.90. 3. The US SEC may propose new crypto regulations as early as this month, providing registration exemptions and safe harbors for startup financing. 4. The crypto fear and greed index fell to 20, indicating the market is back to "extreme fear." 5. $533 million in liquidations occurred across the network in the past 24 hours, with short positions accounting for $294 million. 6. BonkDAO has recovered some of the stolen funds after the governance attack and is investigating the perpetrators. 7. Zcash is nearing completion of the mathematical proof against the hidden issuance vulnerability, and ZEC rose over 12% in a single day. 8. US chip stocks experienced a "stock market crash," with the Fibonacci retracement falling over 6%, making Bitcoin the most resilient asset. 9. The Federal Reserve plans to use the truncated mean instead of PCE to measure inflation, potentially benefiting interest rate cuts this year. 10. SpaceX AI, in partnership with Cursor, unveiled its first AI model on Wednesday, targeting Opus 4.8 and GPT 5.5. 1⃣1⃣ DeepSeek is reportedly developing an AI inference chip to reduce its reliance on Nvidia and Huawei chips, while Zhipu is also reportedly evaluating a self-developed custom AI processor. 1⃣2⃣ Anthropic has secretly filed for an IPO, potentially delaying OpenAI's listing until 2027.
DeepSeek's Secret Self-Developed Inference Chip
According to Beating, Reuters, citing sources familiar with the matter, reports that DeepSeek, a major Chinese model-making company, is developing its own AI inference chip. This research began about a year ago and is primarily for model inference rather than training. In recent months, the company has been privately recruiting chip design engineers. Currently, DeepSeek's model operation and training heavily rely on chips from Nvidia and Huawei. If successful, developing its own inference chip would reduce reliance on external suppliers and provide more cost-effective hardware control. However, the project is still in its early stages and faces manufacturing and memory access restrictions due to US export controls.
Broadcom and Apple extend their custom chip partnership until 2031
PANews reported on July 6th that, according to Bloomberg, Broadcom and Apple have extended their custom chip supply and development partnership until 2031. Building on their existing collaboration on radio frequency and wireless connectivity chips, the two companies will continue their long-term research and development and supply of dedicated chips for products such as the iPhone. The products covered by the agreement will primarily be designed and manufactured in the United States, involving multi-billion dollar long-term procurement commitments.