Meta just unveiled details for their new research supercluster. (RCS) It is a massive supercomputer housed in a data center and when it is fully built-in in 2022, it will be the world’s most powerful supercomputer, And we didn’t know who was going to be The one providing the hardware for the metaverse. Well, now we do. It’s going to be NVIDIA.

This facility for meta Is going to be very impressive and it is going to be full of 760 NVIDIA DGX A-100 systems. It is going to be a complete unit invidious purpose, built for high-end computing in data centers. (Also, going to be cramming 6080 NVIDIA GPUs onto the supercluster, Which is very much Impressive considering the shortage going on right now)

Currently, this RSC delivers a maximum of nearly 1.9 exaflops of performance (Which is very Impressive) However, the second construction of this supercomputer is going to be expanding the RSC to 16000 GPUs and going to be increasing computing power to as much as 5 exaflops.

This is really good news for NVIDIA as they have a very big new customer. And in partnership with Meta, NVIDIA is going to be reaching heights soon. It still is a good thing for them, although they did drop the ARM purchase (Which was a good thing considering we did not want NVIDIA to create a monopoly In the industry)