The leak is somewhat confusing, mainly because it talks about two architectures and many confuse NVIDIA’s decision to stop AMD’s coup with RDNA 3 and return it to them. What is claimed comes from several different sources and together they give a more or less complete picture of what we can expect: NVIDIA Hopper and Ada Lovelace they are not that close.
NVIDIA is again segmenting its architectures and “dividing” them
It’s not something they haven’t done before, but the approach is drastically different, but let’s first place ourselves chronologically and technically before we move on. NVIDIA is go ing to be late in the fight for MCM GPUs, since Intel will hit the first in the games sector, AMD has already done it in the professional sector and the Greens are going to see them coming mainly because it is said and rumor that they encountered problems in the development of the architecture and the switch to Samsung was not as desired.
With that in mind, the rumors are as follows:
- There are two next-generation architectures that will be released at different times with the same architectural base in terms of improvements: NVIDIA Hopper and NVIDIA Ada Lovelace.
- NVIDIA Hopper is likely to be an MCM architecture focused on servers, deep learning, and AI.
- NVIDIA Ada Lovelace is a traditional monolithic architecture that will reach the gaming industry to compete AMD DNA 3.
This is all important because NVIDIA’s strategy is epoch-based. Apparently and from what we saw two weeks ago, the reds and greens are going to use the new TSMC node at 5nm, making them equal in this section despite different approaches.
NVIDIA GPU RTX 40 Hopper and Ada Lovelace, against time?
The problem, as they say, is the weather, as rumors say that by the end of this year at the latest NVIDIA would have Ada Lovelace ready and serial production by TSMC would arrive in the middle of next year. . From what we know, it takes less than a year for NVIDIA to get GPUs into the hands of graphics card makers, so early quotes are for late 2022 for a possible launch.
On the other hand, Hopper could arrive a quarter earlier, in the third quarter, almost a year after AMD itself, but with a completely new architecture and with an internal bandwidth of more than 900 Gb / s, where between the two GPUs we speak of more than 600 Gb / s bidirectional.
NV’s graphics card generally from the output to tape to the market is less than a year, so AD102 is likely to be released from tape in 2022, the earliest may be at the end of this year.
– Greymon55 (@ greymon55) August 14, 2021
So … the NVIDIA RTX 30 will be in another year? Much of it may not be, and there are rumors that NVIDIA has an RTX 30 SUPER ready to fill that gap and keep selling GPUs, while distancing itself from AMD. How much right and wrong is there in this argument? We don’t know, we all find out together little by little, where the only thing that can be confirmed from the mouth of Colette Kress is that Grace Server GPUs would arrive in 2023.
Either NVIDIA has stepped on the accelerator or TSMC is advancing the dates, as the statements predate these new rumors. Strategy to fool AMD perhaps?