Discussion RDNA 5 / UDNA (CDNA Next) speculation

[adroc_thurston]

Looks less like a RDNA4 style tock and more like a clean sheet µarch

1/3/5/7 (every odd tock) is like that.

 

[reaperrr3]

1/3/5/7 (every odd tock) is like that.

That honestly doesn't bode too well, because so far, only the even ones (2 and 4) were actually good.

 

[adroc_thurston]

That honestly doesn't bode too well, because so far, only the even ones (2 and 4) were actually good.

I love superstitions but they just have a roadmap and crank it.
Also RDNA4 sucks balls for obvious reasons.

 

[reaperrr3]

Also RDNA4 sucks balls for obvious reasons.

You mean the lack of halo, or is there something wrong with the uArch itself?
I mean, 64 CUs with GDDR6 virtually matching (and sometimes beating) the 5070Ti is quite alright in my book.
They nearly caught up on a range of issues (V/f, PPW, RT, PT, AI/FSR, mem bw efficiency), and pretty much are slightly ahead on raster IPC for CU vs. SM.

RDNA3 vs. Ada looked FAR worse. Had NV wanted to, they could've priced AMD out of the market then and there.

 

[adroc_thurston]

You mean the lack of halo, or is there something wrong with the uArch itself?

A bit of both.
They cheaped out where they could.

RDNA3 vs. Ada looked FAR worse.

Only in mobile and 3.5 fixed that anyway.

They nearly caught up on a range of issues (V/f, PPW, RT, PT, AI/FSR, mem bw efficiency), and pretty much are slightly ahead on raster IPC for CU vs. SM.

It is indeed a very good shader core; they really haven't shipped any bad ones anyway.