NVIDIA Hopper GPUs Featuring MCM Technology Rumored To Tape Out Soon

Rumor is that NVIDIA's Hopper GPU for the information center marketplace which would exist the beginning to feature MCM technology is taping out soon. This new rumor was posted by Greymon55 in a cryptic tweet but was rapidly decrypted by another Twitter user which 3DCenter caught sight of.

NVIDIA Hopper GPUs For Data Center Taping Out Soon, Volition Feature MCM Technology

The NVIDIA Hopper GPU architecture is presumably designed for the Data Center market. The GPU lineup will be exclusive to the HPC segment, different Ampere which serves both GPGPU (Data Center / Workstation) and gaming markets. The Hopper GPUs are also going to be the offset from NVIDIA to characteristic an MCM or Multi-Chip-Module design which is also beingness utilized by its main competitors, Intel (Xe-HPC) and AMD (CDNA 2) to power their adjacent-generation Ponte Vecchio & Instinct MI200 accelerators.

Expect what we take here ... moving-picture show.twitter.com/ZJmnygr4St

— 3DCenter.org (@3DCenter_org) July 21, 2021

Currently, the NVIDIA roadmap positions Hopper GPU as a 2022 production which means it would launch at least two years after Ampere. It is probable that NVIDIA will denote the GPU around its next GTC upshot in Q2 2022. At present we are hearing reports that NVIDIA is soon going to record out its first Hopper GPUs. The chips are rumored to be based on TSMC's 5nm procedure node and in that location's yet a lot that nosotros don't know virtually the architecture, unlike Intel and AMD who accept their design methodologies exposed through official documents and driver leaks. One time again, the reason for that not happening on NVIDIA'south end is merely considering the GPU isn't even out yet.

Equally for what Hopper is going to offer, Jensen Broune followed upward on another cryptic tweet a few months agone which was posted by Kopite7kimi (a very reliable leaker). The tweet which was a set of numbers is believed to be the cadre configuration of the Hopper GPU. The Hopper GPU could be comprised of ii GPM (Graphics Processing Modules), each of which could feature 8 GPC (4 x two) for a total of 144 SM units.

Hopper is supposed to have 2 of these modules so we are looking at 288 SM units in total. We can't give a rundown on the core count yet since nosotros don't know the number of cores featured in each SMs just if it's going to stick to 64 cores per SM, then we go 18,432 cores which are ii.25x more the full GA100 GPU configuration. NVIDIA could also leverage more FP64, FP16 & Tensor cores within its Hopper GPU which would drive up performance immensely. And that's going to be a necessity to rival Intel'southward Ponte Vecchio which is expected to feature i:ane FP64.

I should explain that.
A GPM of GH100 could have 8(4*2, it seems like GA100)GPC*9Clusters(big changes, add together CPCs).
perf: GH100=3XGA100

— Jensen Bourne (@kopite7kimi) May 25, 2021

Surely, we are still far from an announcement but it's good to know that the Hopper will be taping out presently and we can start hearing more almost it later this yr with an official annunciation next year.

News Source: via Videocardz