Samsung Electronics will use Taiwan Semiconductor Manufacturing Company (TSMC)'s 3-nanometre process to produce base dies for Nvidia's next-generation custom high-bandwidth memory (HBM) chips, according to a report from the Seoul Economic Daily. The report also cites industry sources that Nvidia has asked Samsung, SK Hynix, and Micron Technology to use TSMC's process for NVHBM base dies and sixth-generation DRAM technology for its core dies. Samsung accepted Nvidia's requirements after an internal review but plans…
Key points
Samsung Electronics will use Taiwan Semiconductor Manufacturing Company (TSMC)'s 3-nanome…: Samsung Electronics will use Taiwan Semiconductor Manufacturing Company (TSMC)'s 3-nanometre process to produce base dies for Nvidia's next-generation custom high-bandwidth memory (HBM) chips, according to a report from the Seoul Economic D
The report also cites industry sources that Nvidia has asked Samsung, SK Hynix, and Micro…: The report also cites industry sources that Nvidia has asked Samsung, SK Hynix, and Micron Technology to use TSMC's process for NVHBM base dies and sixth-generation DRAM technology for its core dies.
Samsung accepted Nvidia's requirements after an internal review but plans to retain its i…: Samsung accepted Nvidia's requirements after an internal review but plans to retain its in-house production model for standard HBM and custom products for other customers.
Sources checked by our analysis pipeline: 1
SEOUL, Oct 18 : Samsung Electronics will use TSMC's 3-nanometre process to produce base dies for Nvidia's next-generation custom high-bandwidth memory (HBM) chips, the Seoul Economic Daily reported on Sunday.
The move would mark the first time Samsung has worked with the Taiwanese contract chipmaker at the HBM manufacturing stage, according to the report.
• Nvidia has asked Samsung, SK Hynix and Micron Technology to use TSMC's process for NVHBM base dies, and sixth-generation DRAM technology for its core dies, the newspaper said, citing industry sources.
• NVHBM, unveiled in August, integrates Nvidia's custom memory controller into the base die and is aimed at standardising implementation across multiple memory suppliers, according to the company.
• Samsung accepted Nvidia's requirements after an internal review but plans to retain its in-house production model for standard HBM and custom products for other customers, Seoul Economic Daily said.
• It is also assessing whether its own foundry could produce NVHBM base dies if demand exceeds TSMC's capacity, the newspaper quoted an unnamed source as saying.
The author has decided to archive all their Ruby gems due to a shift in their programming focus, emphasizing the language's future direction towards AI development and the diminishing appeal of contributing to a "commons" stripped by corporations. This decision is a clear indication of a change in their programming interests and a reflection of the language's evolving landscape.
S&P Global predicts that the world's largest tech firms are likely to have more AI-linked debt than previously known, despite the firm not changing any ratings at the moment. The debt issued by AI infrastructure providers and "neocloud" firms, such as Nvidia-backed CoreWeave, has reached increasingly stressed levels due to concerns about the scale of their borrowing. S&P's Global Head for Digital Infrastructure, Mark Habib, highlighted the interconnected financing structures where hyperscalers effectively guarante…
Nvidia introduces two new open models for enterprise and local PC use, aiming to compete with proprietary models and create a new market for its chip-powered PCs. The Nemotron 3.5 Super VL is optimized for enterprise agents, while the Bolt models are lightweight for local use. Nvidia claims faster and more accurate performance compared to competitors, with the Super VL completing tasks faster than Anthropic and OpenAI's models post-trained on specific tasks. Nvidia's vice president of Open Models and Software, Kar…