Image: JDSP

Despite the hopes, dreams, and wishes of green team members worldwide, NVIDIA won’t be tapping TSMC’s 7 nm process for its upcoming lineup of GeForce RTX 30-series graphics cards. That’s according to prominent leaker kopite7kimi, who is “100%” confident about these Ampere-based GPUs leveraging Samsung’s less advanced 8 nm process instead.

Naturally, this can’t be verified until NVIDIA’s official RTX 30-series announcement (which should only be a few months away), but the allegation isn’t all that far fetched based on some of the previous rumors we’ve heard, which include the Korea Herald’s report from 2019 about Samsung’s involvement in Ampere-based GPUs.

“It is meaningful that Samsung Electronics’ 7-nanometer process would be used in manufacturing our next-generation GPU,” said NVIDIA Korea chief Yoo Eung-joon at the time. “Until recently, Samsung has been working really hard to secure (partners) for foundry cooperation.”

The notion of NVIDIA leveraging Samsung’s 8 nm node is also plausible based on recent gossip of RTX 30-series GPUs being power-hungry beasts. And then there’s the fact that TSMC’s 7 nm supply is already extremely tight, thanks to customers such as AMD and Apple.

Join the Conversation

4 Comments

  1. I don’t know, I’m not buying this 8nm "leak." I’d be surprised if Ampere wasn’t on TSMC 7nm, is Samsung’s 8nm even capable of producing a high end chip like GA100? Wouldn’t the die size be even larger, reducing yields vs TSMC? ¯\_(ツ)_/¯
Leave a comment