basix's latest activity
-
Bbasix replied to the thread Discussion AMD’s Custom APU Discussion Thread.For Switch you could optimize your game. For the huge variety of smartphones that will not work. So performance will always be subpar...
-
Bbasix reacted to marees's post in the thread Discussion RDNA 5 / UDNA (CDNA Next) speculation with
Like.
I can think of 2 kinds of cloud hardware 1 identical to AT2 Another based on AT0 but virtualized into 4x (or 3x?) AT3 for a... -
Bbasix replied to the thread Discussion RDNA 5 / UDNA (CDNA Next) speculation.Well, a large customer might be Microsoft ;) But afaik no game streaming service is really killing it. Stadia was cancelled. Nvidias...
-
Bbasix replied to the thread Discussion RDNA 5 / UDNA (CDNA Next) speculation.AT0 might also be used for game streaming services. Maybe also for Xbox game streaming? Pack 6x AT0 together with a 96C Zen 6 EPYC CPU...
-
Bbasix replied to the thread Question Zen 6 Speculation Thread.Well, I have programmed such stuff myself, and parallel I/O requests can speed-up your application significantly. Instead of pushing IO...
-
Bbasix replied to the thread Question Zen 6 Speculation Thread.In some cases, SMT can reach close to 2x performance gains. This can happen in I/O bound situations, e.g. harddisk or networking...
-
Bbasix replied to the thread Discussion RDNA 5 / UDNA (CDNA Next) speculation.Interesting idea. From GPU performance aspects etc. it would match well. But the big advantage of Strix Halo is the big unified LPDDR5...
-
Bbasix replied to the thread Discussion RDNA 5 / UDNA (CDNA Next) speculation.Such a 192 CU chip will definitely be faster than a 5090, when a 5090 is ~1.8x as fast as a 9070 XT with 64CU. Would it be 1.3x / 1.5x /...
-
Bbasix replied to the thread Discussion RDNA 5 / UDNA (CDNA Next) speculation.The slide references to that chip as AT0-XL. When the full design features 192 CU, AMD could go for an AT0-XT with more CUs, full 512bit...
-
Bbasix replied to the thread Discussion RDNA 5 / UDNA (CDNA Next) speculation.No, IF$ cannot reduce the amount of needed VRAM. But if you want a certain amount of VRAM, you need to have a certain bus width. If you...
-
Bbasix replied to the thread Discussion RDNA 5 / UDNA (CDNA Next) speculation."Magnus" could use 384bit + 48 MByte IF$ instead of 96 MByte. In the end a tradeoff between effective bandwidth and amount of VRAM.
-
Bbasix replied to the thread Discussion AMD’s Custom APU Discussion Thread.As far as I know, that one functions completely differently compared to FSR or DLSS. Edit: Here is a presentation from ARM at Siggraph...
-
Bbasix replied to the thread Discussion AMD’s Custom APU Discussion Thread.I just talk about the matrix accelerated part of the algorithm. Everything else would stay on the GPU. That would not add a single frame...
-
Bbasix replied to the thread Discussion AMD’s Custom APU Discussion Thread.Wouldn't that be a really nice use case on APUs? XDNA2 does support FP8 (50 TFLOPS) und is much more powerful than the iGPU in this...
-
Bbasix replied to the thread Discussion RDNA 5 / UDNA (CDNA Next) speculation.It will be far from 2x. But if we see something similar like Turing to Ampere, I would be impressed. Basically 1.3x or so performance...
Facebook
Twitter