• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Discussion RDNA4 + CDNA3 Architectures Thread

Page 303 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

DisEnchantment

Golden Member
1655034287489.png
1655034259690.png

1655034485504.png

With the GFX940 patches in full swing since first week of March, it is looking like MI300 is not far in the distant future!
Usually AMD takes around 3Qs to get the support in LLVM and amdgpu. Lately, since RDNA2 the window they push to add support for new devices is much reduced to prevent leaks.
But looking at the flurry of code in LLVM, it is a lot of commits. Maybe because US Govt is starting to prepare the SW environment for El Capitan (Maybe to avoid slow bring up situation like Frontier for example)

See here for the GFX940 specific commits
Or Phoronix

There is a lot more if you know whom to follow in LLVM review chains (before getting merged to github), but I am not going to link AMD employees.

I am starting to think MI300 will launch around the same time like Hopper probably only a couple of months later!
Although I believe Hopper had problems not having a host CPU capable of doing PCIe 5 in the very near future therefore it might have gotten pushed back a bit until SPR and Genoa arrives later in 2022.
If PVC slips again I believe MI300 could launch before it :grimacing:

This is nuts, MI100/200/300 cadence is impressive.

1655034362046.png

Previous thread on CDNA2 and RDNA3 here

 
Last edited:
While we are waiting, throw us some juicy tidbits 😛
I only know what the leaks say, but think “around 7900XTX” and you’ll be more accurate than he. Shoot, RT is quite a bit faster than RDNA3 by itself, which makes his statement inaccurate. Both IPC and clocks are up as well.
That really sounds like something for the Radeon pro line.
Supposedly it is a gaming SKU.

AMD shot themselves in the foot again. A 96CU 32gb part would’ve really won the day.
 
So in the U.S. maybe $699 and $579 for the 9070XT and 9070 respectively?
That'd be the direct conversion, at least for the cheapest MSRP models.
Seems like AMD and their AIBs might take the same route as Nvidia with magical launch MSRP pricing and everything else has a +$200 availability tax.
 
I assume the 5070 Ti will have a limited supply, just like the 5080, since both use the GB203 chip. Aibs & amd would want to milk before amdiscounts
Yeah, with how scarce graphics cards are right now, they'll sell out at any price point. If AMD or AIBs don't take the profit then scalpers will.
 
It's not possible since the target was cheap GDDR6 and therefore memory controller did not need to spend transistors to support it, plus most likely all GDDR7 is bought up by Nvidia this year.

I'm pretty sure I read here that RDNA4's PHYs and memory controllers support GDDR7, though it hadn't been tested with that memory because there were no SKUs planned with it.
 
Back
Top