The chip race is accelerating: AMD has reached $1 trillion, and China is preparing alternatives to Nvidia

🔼 The chip race is accelerating: AMD has reached $1 trillion, and China is preparing alternatives to Nvidia Demand for computing power for AI continues to accelerate the semiconductor market. We've collected the main news about chips for September. 🟢 AMD reaches $1 trillion capitalization for the first time AMD became the fourth US chipmaker with a capitalization above $1 trillion after Nvidia, Broadcom and Micron. On September 21, the company's shares rose by almost 10%, and since the beginning of the year - by 185%. One of the reasons for the growth is the high demand for AI infrastructure. AMD is accelerating the development of AI chips and is gradually moving from selling individual processors to complex computing systems. 🟢 Alibaba presented its most powerful AI chip Alibaba showed the Zhenwu V900, a new processor for training and running AI models. According to the company, it is three times more powerful than the previous generation and is Alibaba's most powerful AI chip to date. Based on it, the company is preparing its own server platform, and mass release

expected in the first quarter of 2027. Alibaba also plans to increase its data center capacity to more than 20 GW by 2032. 🟢 Huawei has accelerated the release of new chips to compete with Nvidia Huawei has postponed the launch of the next generation of Ascend AI chips to an earlier date. The Ascend 960DT should be released in the first quarter of 2027, and the 960PR version in the third quarter. The company also introduced the new Peerium computing architecture and UnifiedBus technology, which allows you to combine a huge number of processors into a single system. In the future, Huawei expects to scale such systems to 1 million processors. 🟢Qualcomm and Amazon will work together on chips for AI data centers Qualcomm announced a multi-year collaboration with Amazon. Companies will develop custom chips for the AWS infrastructure, which will also focus on inference—the work of already trained AI models. The partners will also work on high-speed optical connections for data transfer within large AI data centers. 📷 [BlueScr

a | Digital Kazakhstan](https://t.me/bluescreenkz)