SK Hynix to Manufacture 3nm HBM4 Chips at TSMC for NVIDIA




/ 7 days ago
SK Hynix to Manufacture 3nm HBM4 Chips at TSMC for NVIDIA

SK Hynix‘s plan to begin manufacturing its next-generation HBM4 memory chips in 2025 was initially centered around using a 5-nanometer process. However, the company now appears to be taking advantage of the strong performance of TSMC’s 3-nanometer process, opting to launch production directly on this more advanced node for its high-end chips.

NVIDIA as the Primary Customer

Nvidia

The main customer for SK Hynix’s HBM4 memory chips will be NVIDIA. Reports suggest that NVIDIA has pushed SK Hynix to expedite its production timeline, with the goal of integrating these chips into its next-generation accelerator GPUs by the latter half of 2025. The rise of artificial intelligence (AI) has positioned NVIDIA as the dominant supplier of AI accelerator cards, and the company is under pressure to stay ahead of the curve as demand for AI hardware continues to surge. This urgency reflects the rapidly evolving nature of the AI market, where companies like NVIDIA must innovate quickly to maintain their competitive edge.

With the leap from a 5-nanometer process to 3 nanometers, SK Hynix expects to see a performance and efficiency boost of approximately 40%.

Collaboration with TSMC and the 3nm Advantage

The decision to use TSMC’s 3-nanometer process is part of a broader collaboration between the two companies, which had already teamed up to develop HBM4 memory technology. By leveraging TSMC’s advanced manufacturing capabilities, SK Hynix can offer a performance advantage over competitors such as Samsung, which is reportedly using a 4-nanometer process for its own HBM4 memory chips.

While SK Hynix is advancing with 3nm production for premium applications, it will also continue to manufacture more conventional HBM4 memory chips at a 12-nanometer process. This bifurcated approach allows SK Hynix to meet a wider range of customer needs, reserving the advanced 3nm process for high-end applications like NVIDIA’s AI accelerators, while the 12nm chips cater to more standard use cases.


Topics: ,

Support eTeknix.com

By supporting eTeknix, you help us grow and continue to bring you the latest newsreviews, and competitions. Follow us on FacebookTwitter and Instagram to keep up with the latest technology news, reviews and more. Share your favourite articles, chat with the team and more. Also check out eTeknix YouTube, where you'll find our latest video reviews, event coverage and features in 4K!

Looking for more exciting features on the latest technology? Check out our What We Know So Far section or our Fun Reads for some interesting original features.

eTeknix Facebook eTeknix Twitter eTeknix Instagram eTeknix Instagram
  • Be Social With eTeknix

    Facebook Twitter YouTube Instagram Reddit RSS Discord Patreon TikTok Twitch
  • Features


Send this to a friend
})