NVIDIA’s next-generation GeForce RTX 40 sequence gaming graphics playing cards primarily based on the Ada Lovelace GPU structure are being prepped for a serious 2022 launch. In its newest report, DigiTimes states that companions together with Taiwan factories which are companions with NVIDIA are preparing for a serious GPU refresh subsequent 12 months within the type of the GeForce RTX 40 sequence.
NVIDIA Companions at Taiwan Prep For Main GeForce RTX 40 ‘Ada Lovelace’ Sequence GPU Launch in 2022, Gaming GPUs To Make the most of TSMC’s 5nm Course of Node
Now we have already heard about the potential of NVIDIA using TSMC’s 5nm course of node for its next-generation gaming GPUs codenamed Ada Lovelace from dependable leakers however this time, the data comes from straight throughout the Taiwanese primarily based factories the place these GPUs can be made. Whereas the DigiTimes article is behind a paywall, a snippet of the data was revealed by RetiredEnginner (@chiakokhua) over at Twitter.
“Nvidia’s biennial GPU refresh coming in 2022, driving on metaverse and gaming. Following H100, primarily based on Hopper structure, utilizing TSMC’s 5nm + CoWoS, geared toward datacenter/AI, gaming GPU RTX40 sequence, primarily based on Ada Lovelace structure, can even faucet TSMC’s 5nm….”
— RetiredEngineer® (@chiakokhua) November 30, 2022
The NVIDIA Ada Lovelace GPUs will energy the next-generation GeForce RTX 40 graphics playing cards that may go head-on with AMD’s RDNA 3 primarily based Radeon RX 7000 sequence graphics playing cards. There’s nonetheless some hypothesis concerning the usage of MCM by NVIDIA. The Hopper GPU, which is primarily aimed on the Datacenter & AI section, is allegedly taping out quickly and can characteristic an MCM CoWoS structure. NVIDIA will not be utilizing an MCM design on its Ada Lovelace GPUs so they may preserve the normal monolithic design. The Ada Lovelace GPUs are anticipated to usher in a sequence of key improvements, architecturally.
NVIDIA GeForce RTX 4090 Graphics Card – Ada Lovelace Powered AD102 Flagship GPU
Primarily based upon earlier rumors, there have been whispers that NVIDIA would make the most of TSMC’s N5 (5nm) course of node for its Ada Lovelace GPUs. This consists of the AD102 SKU too which can be a wholly monolithic design. In his newest tweet which talks in regards to the particular GPU configurations, the AD102 GPU is alleged to characteristic a clock pace as excessive as 2.5 GHz (2.3 GHz common enhance). The particular tweet states that the GPU clock for Ada Lovelace ‘AD102’ might be 2.3 GHz or larger so let’s take that as a baseline and beforehand leaked specs to determine the place the efficiency ought to land.
The NVIDIA AD102 “ADA GPU” seems to have 18432 CUDA Cores primarily based on the preliminary specs (which might change), housed inside 144 SM models. That is virtually twice the cores current in Ampere which was already an enormous step up from Turing. A 2.3-2.5 GHz clock pace would give us as much as 85 to 92 TFLOPs of compute efficiency (FP32). That is greater than twice the FP32 efficiency of the present RTX 3090 which packs 36 TFLOPs of FP32 compute energy.
The 150% efficiency leap appears large however one ought to keep in mind that NVIDIA already gave an enormous leap in FP32 numbers this era with Ampere. The Ampere GA102 GPU (RTX 3090) gives 36 TFLOPs whereas the Turing TU102 GPU (RTX 2080 Ti) provided 13 TFLOPs. That is over a 150% improve in FP32 Flops however the real-world gaming efficiency improve for the RTX 3090 averaged at round 50-60% quicker over the RTX 2080 Ti. So one factor we should not overlook is that Flops do not equal GPU gaming efficiency lately. Moreover, we do not know if 2.3-2.5 GHz is the typical enhance or the height enhance with the previous which means that there might be even increased compute potential for AD102.
Apart from that, the leaker additionally states that the NVIDIA GeForce RTX 40 flagship would retain a 384-bit bus interface, much like the RTX 3090. What’s fascinating is although that the leaker mentions G6X which implies that NVIDIA will not be shifting to a brand new reminiscence commonplace till after Ada Lovelace and make the most of the upper pin-speeds of G6X of 21 Gbps for its next-generation playing cards earlier than we see a more recent commonplace (e.g. GDDR7). The cardboard will characteristic 24 GB of reminiscence so we will both anticipate single-sided 16Gb DRAM or dual-sided 8Gb DRAM modules.
NVIDIA CUDA GPU (RUMORED) Preliminary:
| GPU | TU102 | GA102 | AD102 |
|---|---|---|---|
| Structure | Turing | Ampere | Ada Lovelace |
| Course of | TSMC 12nm NFF | Samsung 8nm | 5nm |
| Graphics Processing Clusters (GPC) | 6 | 7 | 12 |
| Texture Processing Clusters (TPC) | 36 | 42 | 72 |
| Streaming Multiprocessors (SM) | 72 | 84 | 144 |
| CUDA Cores | 4608 | 10752 | 18432 |
| Theoretical TFLOPs | 16.1 | 37.6 | ~90 TFLOPs? |
| Reminiscence Sort | GDDR6 | GDDR6X | GDDR6X |
| Reminiscence Bus | 384-bit | 384-bit | 384-bit |
| Reminiscence Capability | 11 GB (2080 Ti) | 24 GB (3090) | 24 GB (4090?) |
| Flagship SKU | RTX 2080 Ti | RTX 3090 | RTX 4090? |
| TGP | 250W | 350W | 450-650W? |
| Launch | Sep. 2018 | Sept. 20 | 2022 (TBC) |
The NVIDIA Ada Lovelace GPUs will energy the next-generation GeForce RTX 40 graphics playing cards that may go head-on with AMD’s RDNA 3 primarily based Radeon RX 7000 sequence graphics playing cards. There’s nonetheless some hypothesis concerning the usage of MCM by NVIDIA. The Hopper GPU, which is primarily aimed on the Datacenter & AI section, is allegedly taping out quickly and can characteristic an MCM structure. NVIDIA will not be utilizing an MCM design on its Ada Lovelace GPUs so they may preserve the normal monolithic design.
