AMD Vega 10, Vega 20, Vega 11 and Navi 110 GPU Roadmap Leaked

Navi was originally scheduled for release in 2022.

AMD's major server roadmap, detailing their side by side generation Vega and Navi GPUs has been leaked past Videocardz. The roadmap that spans upward to 2022 shows how AMD will configure their next-gen GPUs into unlike server segments that include applications similar HPC, Deep Learning and Inferencing.

AMD Vega and Navi GPU Roadmap For Servers Leaked - Vega Arrives in 2022, Navi in 2022

The latest roadmap leak comes merely a day after major specifications were likewise revealed by Videocardz. 2022 started off with AMD replacing most of their graphics lineup with Polaris based cards. With the exception of loftier-end products, the Polaris 10 and Polaris 11 cards based on FinFET architecture delivered skilful improvements beyond the lath. Whether that be in terms of efficiency or operation but the most stellar selling point for these cards was their competitive pricing.

Moving on, AMD is expected to launch new high-stop, enthusiast products in the market place. But the products nosotros are looking at today are not aimed at consumers or gaming enthusiasts but rather, the server market. And so let's have a look what 2022 and beyond has to offer in terms of server aimed products from AMD.

AMD Radeon Pro Serial To Go Radeon Pro S9 Nano

The AMD Radeon Pro serial was introduced two months back and featured a range of Polaris ten, Polaris xi and Fiji based cards. The primary attractions were the Radeon Pro WX 7100 and Radeon Pro SSG. The old is the near affordable work-station card while the latter is a heavy-duty, Fiji based card that comes with a on-board SSD solution to deliver immense boost in productivity and performance on workstation suites. More on these ii cards hither.

AMD plans to introduce a new server-focused menu in this lineup besides. Based on their Fiji GPU, the card would be known as Radeon Pro S9 Nano. The "WX" stands for workstation while "Southward" stands for server. The solution is expected to characteristic the same specifications as the Radeon R9 Nano forth with some clock adjustments. The new launch could happen in a matter of weeks equally mentioned in the leak. Radeon Pro S9 Nano will aim the inference market and can be sighted as competition against NVIDIA's recently launched Tesla P40 solution for inferencing.

The R9 Nano, although expensive, was an impressive production with insane specs in a SFF solution.

Of course, the Radeon Pro S9 Nano won't be the only inferencing card available. AMD volition also launch cheaper cards based on Polaris 10 and Polaris eleven GPUs. These will go in nicely against the Tesla P4 solution with peradventure affordable price bracket.

AMD Vega x, Vega 11, Vega 10 x2 and Vega 20 In 2022 and 2022

In 2022, AMD finally plans to upgrade all of their GPUs to the new FinFET compages. While Polaris serves the mainstream, Vega will be aiming for the high-end marketplace. The new graphics chip for Deep Learning will be the Vega 10, a very beefy GPU which we detailed yesterday. AMD Vega x is expected to launch in beginning quarter of 2022.

From the leaked details, we now know that Vega 10 can feature up to 4096 stream processors based on a new GCN uARCH (GFX9). The chip has every bit much every bit 24 TFLOPs of one-half-precision and 12 TFLOPs of single precision compute operation. The chip also features up to xvi GB of HBM2 that is clocked to operate at 512 GB/s. The cards based on this chip are rated with 225W TDP.

AMD's Vega lineup will feature immensely powerful cards with HBM2 memory in 2022. Original Epitome by FallenZeraphine

Now nosotros noticed that Vega 10 isn't aiming the HPC market just instead, focused at the Deep Learning servers. It will have two parts bachelor in 2022, a single chip and dual chip solution. Both feature double precision at 1/16th the rate of FP32. Information technology confirms that Vega 10 is similar to GP102 which cuts downwards double precision compute from the more than beefy GP100 GPU. The details on this will be covered in a more informative commodity later on.

AMD Vega 20 Will Be Their First FinFET Based HPC Chip - Launching in 2H 2022

This shows that Hawaii volition still stand up as the HPC muscle up till 2022. A new GPU will be adult past AMD in the meanwhile known as Vega twenty with DP rated at 1/ii of SP. This would be a solution targeted at GP100 or the new GV100 scrap based on Volta which is expected to be introduced past then.

The latter is more possible since Vega xx would not only utilize a enhanced congenital design merely also come packaged in the new GloFo 7nm FF process. Global Foundries has already announced take chances product of 7nm FF in 2022 so nosotros might only go to see Vega xx in some form in that year. With upward to 32 GB of HBM2 memory (i TB/s) and a fast 64 compute unit design, this chip will have impressive double precision capabilities for HPC platforms.

Also, do yous remember that we mentioned a high-functioning interconnect beingness developed by AMD a while ago? Looks like the solution will be known as xGMI. This peer-to-peer solution volition be embedded in HPC servers that include Vega twenty and Zen based Naples processors.

AMD Next Generation Vega x, 11, xx and Dual GPU Graphics Card Rumored Lineup:

WCCFTech	Polaris 10	Vega 11	Vega 10	Vega 10 Dual GPU	Vega 20
Twelvemonth	2016	2017	2017	2017	2018
Process	14nm FinFET	14nm FinFET	14nm FinFET	14nm FinFET	7nm FinFET
Transistors In Billions	five.7	TBA	TBA	TBA	TBA
Stream Processors	2304	2304+ (est.)	4096	8192	4096
Clock Speed	1266 Mhz	TBA	1526 Mhz	1100 Mhz+ (est.)	1800 Mhz+ (est.)
Performance	5.8 TFLOPS	TBA	12.5 TFLOPS	19 TFLOPS - 24 TFLOPs (est.)	15 TFLOPS+
TDP	150W	TBA	225W	300W	150-300W
Memory	8GB GDDR5	TBA	8GB/16GB HBM2	sixteen-32GB HBM2	16-32GB HBM2
Memory Omnibus	256bit	TBA	2048-bit (2 Stacks)	4096-scrap (2048-bit x2)	4096-bit (4 Stacks)
PCI Limited	iii.0	TBA	three.0	3.0	4.0
Bandwidth	256 GB/s	TBA	512 GB/southward	i TB/south	1 TB/s

AMD Navi 10, Navi 11, Navi ten x2 in 2022, Navi 20 in 2022?

Moving forrad, AMD has plans to ship Navi in 2022. Now you lot might believe that Navi has been delayed but nosotros don't find that as the example. This roadmap is specific to the server platform and these products accept the tendency to arrive later as compared to consumer parts. Hence, Navi could however be a planned 2022 launch on the consumer side but could miss that time frame on the server side of things.

Next-Gen GPUs will leverage from more than advanced retentivity solutions such as HBM3.

However, Navi x and dual Navi cards are expected to transport in 2022. Nosotros tin expect these cards to leverage from new GCN architecture, ane that is revised extensively for greater efficiency spring over Vega. There's as well next-gen memory (HBM3) and PCI-e 4.0 in the talks. Navi 10 will be aiming the Deep Learning marketplace while the smaller Navi eleven core will be aimed at the Inference market.

Only similar Vega, Navi won't see a HPC solution for a while. The Vega 20 GPU volition be remain at the helm of AMD's HPC offerings and eventually be replaced past a powerful double Precision part in the form of Navi xx. This solution tin can striking the market effectually 2022. That's information technology equally far as AMD's server roadmap is concerned, we will encompass some specifications comparison betwixt NVIDIA Pascal/Volta and AMD Vega/Navi GPUs in the days alee to assess their potential in the HPC markets.

AMD Server GPU Roadmap 2022-2019

	2016	2017	2018	2019
Production Segement
HPC	Hawaii	Hawaii	Vega twenty	Vega 20
Deep Learning	Polaris ten Polaris 11	Vega 10 Vega 10 x2	Vega x Vega 10 x2	Navi ten Navi 10 x2
Inference	Radeon Pro S9 Nano	Vega 11	Vega 11	Navi 11