Citing surging AI demand and tight semiconductor supply chain capacity, TSMC is reportedly considering a 10%–20% price increase for its CoWoS advanced packaging services.
SambaNova announced on October 27 its collaboration with Oracle to bring its SN50 RDU AI inference chip onto OCI. Oracle is building a diverse AI compute portfolio that includes NVIDIA, AMD, Positron Atlas, Google TPU, Cerebras wafer-scale chips, and Intel Xeon Diamond Rapids.
SK Hynix reported third-quarter revenue of 104.62 trillion won, up 327.9% year on year, and operating profit of 83.33 trillion won, surging 632.1%, with quarterly revenue exceeding 100 trillion won for the first time. The company expects HBM4 shipments and revenue to surpass HBM3E in the fourth quarter and said memory shortages will persist until 2030.
ByteDance, Alibaba and Tencent have shown combined demand for more than 3 million Nvidia RTX PRO 5500 GPUs, representing a potential order value of over $35 billion, though Nvidia may not be able to fulfill all requests and actual deliveries could be much lower.
OpenAI reportedly plans to purchase Google's AI chips, with financial terms and other details of the deal still unclear.
Oracle announced it is expanding its cloud AI chip agreement with Nvidia to multi-gigawatt scale for Nvidia's latest Vera Rubin AI servers, and will offer Google TPU chips for the first time, while planning to add inference chips from SambaNova and Cerebras to diversify its chip lineup.
Teradyne announced the launch of the UltraFLEXplus S-Series semiconductor manufacturing equipment platform, an extension of the UltraFLEXplus platform designed for edge AI applications. The first model, UltraFLEXplus S24, is a 24-slot system targeting IDM and OSAT manufacturers.
On October 27, Apple released the 14-inch MacBook Pro and iMac equipped with the M6 chip, starting at 15,999 yuan and 12,499 yuan. Pre-orders open at 1:10 a.m. on October 29, and official sales begin on November 6.
On Oct. 26, Cadence Design Systems introduced AI models optimized for computer chip design that chipmakers can run internally, deepening its open-source model cooperation with Nvidia and distinguishing it from Synopsys' recent agreement with OpenAI.
According to supply chain sources, ChangXin Memory recently declined to quote Apple for DRAM supply, prioritizing the Chinese domestic market. In August, Apple's price-cut request for LPDDR5X and other mobile DRAM had been rejected. ChangXin's capacity is fully booked by long-term orders from domestic terminal makers.