Now that AMD is concluded launching its RDNA2 household of GPUs, speculation about what it has in keep for subsequent-gen is ramping up. As with all rumors this new data has to be taken with a grain of salt, but the latest details are juicy indeed. The huge information is AMD is increasing both of those the amount of money of memory on its flagship GPU and memory bandwidth far too. And that’s just before we even get started discussing its monstrous amount of Infinity Cache.
This round of rumors arrives from alleged insider @Greymon55 on Twitter. This time all over they are tipping the memory bus width for AMD’s upcoming GPUs. They are acknowledged as Navi 31, 32, and 33, and go in descending get with Navi 31 becoming the RX 7900 XT. The leading-shelf chip will reportedly be rocking an further-broad (for AMD) 384-bit memory bus, which is a important enhance from the RX 6900 XT’s 256-bit bus. In the present-day technology of GPUs only Nvidia uses that vast of a memory bus. It takes advantage of it on its RTX 3080 12GB cards and those people previously mentioned it in the product stack. AMD has been able to get by with a narrower 256-bit bus because it has Infinity Cache to increase it.
Offered the use of a broader memory bus, it’s also rumored it will be outfitting it with 24GB of GDDR6. This is an boost from 16GB in the recent RX 6900/50 XT. This could assistance AMD contend against the RTX 4090, and it would announce AMD is carried out actively playing close to. It by no means matched Nvidia’s RTX 3090/Ti 24GB, but it appears to be like to be intent on taking up that combat with RDNA3.
In addition to a broader memory bus, and much more memory, AMD is also raising Navi 31’s Infinity Cache also. As revealed in the mockup higher than, it is considered AMD will include 64MB of infinity cache on each and every of the package’s 6 4GB memory modules. No matter whether this is 3D stacked like on the Ryzen 7 5800X3D or not remains to be witnessed. This is assuming the package has seven chiplets as previously claimed. This would grant the GPU a surprising 384MB of L3 cache. That is a large update more than to 128MB in the RX 6950 XT. This would make the Navi 31 a 4K gaming beast, with sufficient memory bandwidth for significant-resolution, ray-traced gaming.
Of training course, there is continue to a good deal of speculation about whether or not this 6+1 chiplet layout will be the ultimate Navi 31 package deal. Rumors have various hence considerably about the layout of its seven chiplets. What is crystal clear is AMD has taken a groundbreaking strategy for Navi 31 vs . an evolutionary tactic like Nvidia will be performing with Ada Lovelace. Nvidia is sticking with a monolithic die this time about. AMD is having a various path with a multi-chip module (MCM) tactic. Which style will finally verify outstanding is one thing we’re all very curious to find out.
Unfortunately, it might be awhile till all the playing cards are on the desk. While Nvidia is tipped to launch its RTX 40-sequence in mid-July, AMD is not envisioned to launch RDNA3 any time soon. Rumors point out it’ll probably start close to its Zen 4 CPUs, which are owing in the latter part of the 12 months.