**Nvidia’s Next-Gen AI Superchip Platform is Finally Here (Kind of)**
CES 2023 has come and gone, and Nvidia’s been making some major waves with their announcements. As I’m sure you’re aware, the company’s next-gen AI superchip platform, Vera Rubin, has been a long time coming. And the good news is that it’s finally heading to clients in 2023 – better late than never, right?
So, what’s the big deal about Vera Rubin? For starters, it’s going to reduce the cost of running AI models by a whopping 90% compared to Nvidia’s current top-of-the-line chip system, Blackwell. That’s like going from a gas guzzler to a hybrid – game. changer. And if that’s not enough, Rubin can train large AI models using a whopping 75% fewer chips than Blackwell. I mean, who wouldn’t want to save that kind of cash and resources?
But it’s not just about the specs (although, let’s be real, they’re pretty impressive). Nvidia’s already got some big-name partners on board, including Microsoft and CoreWeave, who are already offering services powered by Rubin chips later this year. And we’re talking thousands of chips being installed in two massive AI data centers in Georgia and Wisconsin – that’s a big deal.
So, what does this mean for the future of AI? It means that Nvidia’s clients will have access to a more affordable and efficient way to train and run their AI models, and that’s a huge deal. And let’s not forget that Nvidia’s teaming up with Red Hat to offer more products that can run on the Rubin chip system, which will bring even more flexibility to businesses and organizations looking to implement AI.
The chip platform itself is named after Vera Rubin, an American astronomer who revolutionized our understanding of galaxies. And when you see the specs, you can understand why – the Rubin chip system includes six different chips, including the Rubin GPU and a Vera CPU, both of which are built using the latest 3-nanometer fabrication process and top-of-the-line bandwidth memory technology. It’s like a well-oiled machine, folks.
So, what does “full production” mean for Vera Rubin? Nvidia’s been saying that systems built on Rubin would start arriving in the second half of 2026, but it looks like they’re ahead of schedule. We can expect to see the first Rubin chips hitting the market later this year, and I’ll be keeping an eye on the story as more information becomes available. Stay tuned!
—
[ad]
