Samsung and Nvidia are making an AI megafactory powered by 50,000 GPUs

New Delhi: Samsung have announced their plans to establish a next-generation artificial intelligence (AI) “megafactory” in partnership with the US chipmaker Nvidia. The initiative, revealed on Friday and reported first by Focus Taiwan, aims to integrate AI throughout Samsung’s entire semiconductor production ecosystem. 

Korean semiconductor Samsung noted that they are planning to operate on more than 50,000 Nvidia GPUs and serve as an “intelligent manufacturing platform” that will be capable of predicting, analysing and optimising chip production in real time. 

Also Read

Google denies claims of using Gmail messages to train AI as lawsuit emerges

Voter list update scam explodes nationwide—Here’s what you must never share

X rolls out new transparency tool showing users’ country of origin

The brand said, “The Samsung AI Factory goes beyond traditional automation.” What it will do is connect and interpret the immense data that is generated across chip design, production and equipment operations as well. 

For Nvidia, this is their latest big partnership, with chips remaining the essential aspect for making and deploying advanced AI. The deal with Samsung comes right after Nvidia CEO Jensen Huang's announcement in Washington, D.C. on Tuesday that the brand was collaborating with Palantir, Eli Lilly, CrowdStrike, and Uber.

Raymond Teh, Nvidia’s senior vice president of Asia-Pacific, on Wednesday told media that they were working closely Korean government in order to support their big plans for AI. The partnership supports Huang’s claim that on Tuesday that Nvidia has a book of business that amounts to $500 billion from their present-gen GPU known as Blackweel, along with next-gen GPU, the Rubin. 

The forecast helped the brand become the first to reach the market cap of $5 trillion.

Representatives of Nvidia have said that they will collaborate with Samsung in order to modify the Korean brand’s chipmaking lithography platform for use with Nvidia's GPUs. They further said that the process will result in 20 times better performance for Samsung. Samsung will also make use of Omniverse, which is a simulation software developed by Nvidia. 

Samsung, quite well-known for its smartphones, recently also announced that they would run their own AI models on Nvidia chips. Samsung is an important supplier to Nvidia in apart to being a customer and partner.

Along with their AI chips, Samsung produces high-performance memory that Nvidia employs extensively. Samsung announced that they will also collaborate with Nvidia to tweak the HBM4 memory for use in AI chips.