Toronto-based chip designer Tenstorrent taps Samsung Foundry US to build its new generation of artificial intelligence (AI) chips. Samsung’s Foundry Design Service team will develop the chips using its new SF4X process.
In addition to AI chips, Samsung will also build Tenstorrent’s central processing units (CPUs). Both product categories are designed according to Tenstorrent’s new RISC-V chip design.
Samsung US Business Boosted With New Chip
Samsung will make Tenstorrent’s chips at its factory in Austin, Texas, where the firm is presently erecting another factory to expand US production. Tenstorrent’s CEO Jim Keller said Samsung’s expertise in electronics manufacturing will help it bring Tenstorrent’s chips to consumers.
“Samsung Foundry’s commitment to advancing semiconductor technology aligns with our vision for advancing RISC-V and AI and makes them an ideal partner to bring our AI chiplets to market.”
Tenstorrent’s choice of Samsung is a predictable move for a member of the struggling US chip industry. Samsung’s US business can make chips according to a four-nanometer (4nm) process, which some view as close to state-of-the-art.
Its new facility in Austin aims eventually to make three-nanometer (3nm) chips which are 30 percent more powerful but consume half the power. But Samsung is not expected to start mass-manufacturing three-nanometer chips until 2024, giving its Asian competitor, Taiwan Semiconductor Manufacturing Company (TSMC), a chance to pull ahead in the race.
TSMC already manufactures 3nm chips for the A17 chip in Apple’s latest iPhones and has dominated semiconductor manufacturing since the US fell behind several years ago. President Joe Biden has supported the construction of several new semiconductor manufacturing facilities in the United States.
How AI Is Changing Approaches to Chip Design
The global war for semiconductor supremacy has come into sharp focus with the advent of AI. Advances in large language models, such as those used by OpenAI and Google, are directly tied to the capabilities of semiconductors.
As a result, several companies, such as Tenstorrent, Nvidia, and Graphcore, have invested significantly into maximizing the ability of semiconductors to process AI models. Tenstorrent uses RISC-V, an open-source set of instructions that chip designers can rearrange to optimize calculations for AI.
Read more: 9 Best Artificial Intelligence Stocks To Buy in 2023
Graphcore, a UK-based company backed by Microsoft and SoftBank, has developed a chip that can work on multiple streams of data at the same time. Nvidia’s software tools can customize its Graphics Processing Units for large-scale AI workloads in big data centers.
Read more: How Will Artificial Intelligence (AI) Transform Crypto?
Got something to say about the deal between Tenstorrent and Samsung to build AI chips in the US or anything else? Write to us or join the discussion on our Telegram channel. You can also catch us on TikTok, Facebook, or X (Twitter).
Disclaimer
In adherence to the Trust Project guidelines, BeInCrypto is committed to unbiased, transparent reporting. This news article aims to provide accurate, timely information. However, readers are advised to verify facts independently and consult with a professional before making any decisions based on this content.
Be the first to comment