We’re looking forward to Nvidia’s next generation of graphics cards, and it looks like Samsung has big plans to be a part of that, or at least they aim to. They lost out to TSMC when it comes to producing chips integral to the current 40 series, and TSMC has been on top ever since. However, an internal Samsung strategy looks to capitalize on their development of more the efficient 3nm GAA process, with aims to mass produce chips within the first half of 2024.
Samsung Foundry prioritizes 3nm chips
According to a report from Korean news outlet, FN news, Samsung is looking to prioritize orders from Nvidia by mass producing 3nm chips, down from the 5nm process size provided by TSMC that helps power their Ada Lovelace cards, the RTX 40 series. Not just that though, as the recent Hopper and Blackwell families are also using TSMC’s 5nm process, giving Samsung the incentive to double down on 3nm in order to fight off TSMC.
We previously saw Samsung’s 8nm process nodes used for the RTX 30 series, based on Ampere architecture. It seems that ever since Nvidia went on to adopt TSMC’s 5nm process for the RTX 40 series, Samsung has been looking to gain back their partnership with the GPU giant. And the more efficient 3nm is the way to do that. According to a report from Wccftech, Samsung has been pushing this technology as much as possible and has seen some breakthroughs.
Will Nvidia stop using TSMC chips?
We’re yet to know which process node Nvidia will be utilizing on the upcoming RTX 50 series, or beyond. TSMC still has the leading position as of now, but these developments from Samsung could held turn the tide in their favor.
Wccftech reports that Samsung has a strategy in place, codenamed ‘Nemo’, which refers to their effort on receiving orders from Nvidia for these new chips. At the moment, it does seem like there is some time to go until Nvidia makes a full move back to Samsung, and we’re looking forward to the Computex and Hot Chips 2024 to see what’s in store.
