AMD Radeon RX 7900 XT ‘Big Navi 31’ RDNA 3 GPU To Feature Up To 60 WGPs & Up To 15,360 Cores

AMD Radeon RX 7900

A model new rumor concerning AMD’s next-generation and flagship RDNA 3 GPU, the Huge Navi 31, which goes to energy the Radeon RX 7900 XT graphics card, has been printed by Beyond3D forums (through 3DCenter). The rumor means that AMD is dropping a very fashionable GPU terminology from its RDNA 3 lineup.

AMD Radeon RX 7900 XT ‘Huge Navi 31’ RDNA 3 GPU To Characteristic Up To 60 WGPs For A Whole of 15,360 Cores

We’re seeing loads of rumors concerning AMD’s RDNA 3 GPUs pop up not too long ago. Finally, we heard, the AMD Huge Navi GPU inside the RDNA 3 lineup, the Navi 31, was going to function 240 Compute Items or 120 CUs per die for a complete 15,360 cores. However, Beyond3D discussion board member, Bondrewd, is sort of assumed that the period of the CU or Compute Unit is over and AMD is transferring over to WGP or Work Group Processors as the primary core block of its next-generation GPUs.

AMD Radeon RX 6600 XT RDNA 2 Graphics Card To Be Unveiled at Chinajoy 2021 This Weekend

The Compute Unit has been the basic block of the AMD GPU household for nearly a decade however with next-generation MCD (Multi-Chiplet Die) GPUs, the general variety of CUs goes to extend dramatically. As such a dramatic change can also be required and AMD might need to determine whether to make use of WGP as the primary compute block. As reported beforehand, every Navi 31 GPU die will function with 30 WGPs and 256 stream processors per WGP. This can give 7,680 cores per die and 15,360 cores in whole.


Additionally what’s attention-grabbing is that each KittyYYuko & Kopite7kimi state that AMD RDNA 3 ‘Huge Navi 31’ flagship might function 60 WGPs which equals 120 Compute Items. There’s additionally hypothesis concerning a really totally different CU design for chiplet primarily based GPUs and the usual monolithic choices corresponding to Navi 33. There are additionally important enhancements to the front-end shader engine.


— Yuko Yoshida (@KittyYYuko) July 26, 2021


— kopite7kimi (@kopite7kimi) July 22, 2021

ASRock X570S PG Riptide Motherboard Assessment, Can This $185 US Board Deal with A $799 US AMD Ryzen 9 5950X CPU?

The reminiscence design will get extra attention-grabbing as AMD RDNA 3 GPUs are going to go insane on the Infinity Cache resolution. The Navi 31 ‘Radeon RX’ GPUs are anticipated to function as much as 512 MB of Infinity Cache, virtually quadruple the quantity featured on Navi 21 however retain the 256-bit bus interface. The playing cards are anticipated to retain GDDR6 reminiscence dies however with quicker & optimized frequency/energy specs.

If the Navi31’s cache reaches 512MB, that does clarify why the bitwidth stays at 256bit.

— Greymon55 (@greymon55) July 16, 2021

256bit GDDR6
256/512mb IC

— Greymon55 (@greymon55) July 22, 2021

There have been a number of rumors stating that the upcoming RDNA 3 GPUs are going to outperform no matter what NVIDIA has to supply by way of rasterization efficiency. It appears to be like like AMD will take the lead by providing the primary MCM-powered GPUs beneath its Radeon RX graphics card lineup. However, at the identical time, NVIDIA is anticipated to rapidly transition to its MCM GPU lineup which is able to supply over 3x efficiency increase over Ampere GPUs. Following is the anticipated efficiency improvement:

AMD RDNA GPU (Generational Comparability) Preliminary:

GPU IdentifyNavi 10Navi 21Navi 31
GPU Course of7nm7nm5nm (6nm?)
GPU Package dealMonolithicMonolithicMCD (Multi-Chiplet Die)
Shader Engines246
GPU WGPs204030 (Per MCD)
60 (In Whole)
SPs Per WGP128128256
Compute Items (Per Die)4080120 (per MCD)
240 (in whole)
Cores (Per Die)256051207680
Cores (Whole)2560512015360 (2 x MCD)
Reminiscence Bus256-bit256-bit256-bit
Reminiscence KindGDDR6GDDR6GDDR6
Infinity CacheN/A128 MB256-512 MB
Flagship SKURadeon RX 5700 XTRadeon RX 6900 XTRadeon RX 7900 XT
LaunchQ3 2019This autumn 2020This autumn 2022

The AMD RDNA 3 ‘Navi 3X’ GPU-powered Radeon RX graphics card lineup is anticipated to supply as much as 3x efficiency enhancement over current RDNA 2 choices. It could mark a significant leap ahead for graphics evolution and now that AMD is already onboard with next-gen applied sciences corresponding to FSR and Raytracing, anticipate really heated next-gen competitors between the pink and inexperienced groups.

Leave a Reply

Your email address will not be published. Required fields are marked *