Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)

Page 756 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Geddagod

Golden Member
Dec 28, 2021
1,683
1,720
136
People also underestimate the usefulness of AVX-512. While it is not the easiest to add in support for, it provides fantastic speedups for the relevant workloads. With Zen 4 and Zen 5 becoming mainstream, developers have more incentive to add support for it. Especially as Intel continues to falter.
People overestimate the usefulness of AVX-512. The vast, vast, majority of consumers are never going to interact with it or care. Also, Intel isn't faltering that much in the client market...
Gamers Nexus video addressing all the most common claims around “Reviewers do not know how to test”

Still, pretty mid testing. Deep dive? Bruh, this could be so much deeper. I mean, it is a good video in general, but again, there is so much more testing they could have done, not with just different applications, but digging deeper into the power curves, and what parts of the CPU consume more power.
TSMC s N4P provide 28% better perf/watt than their regular N5 but it happen than Zen 4 use N5P wich has 11% better perf/Watt than the vanilla N5.

Hence the perf/Watt improvement at isofrequency brought by N4P is only 1.28/1.11 = 15.3% better perf/Watt or 5.7% better perf at isopower.

I love how you have to keep changing the numbers and what node AMD uses in order to get the numbers to fit lmao
 

mostwanted002

Member
Jun 16, 2023
63
127
86
mostwanted002.page

New AGESA in BETA BIOS for ASUS X670E boards.
 

Josh128

Banned
Oct 14, 2022
1,542
2,295
106
Looking forward to sipping some coffee whilst pouring over some 9950X PBO benchmarks this morning. A small joy in my otherwise stressful day.

Also, to kick a less-kicked dead horse, do we have any reason to believe that 9800X3D will have any more % gains in gaming over 7800X3D than 9700X has over 7700X, other than the potential extra 3%-6% tops (300Mhz) frequency they might have been able to eek out of the X3D this go round? Like, people are saying Zen 5 is mem bandwidth constrained-- any chance X3D L3 will allow the new arch to stretch its legs more than it did Zen 4?
 

Josh128

Banned
Oct 14, 2022
1,542
2,295
106

misuspita

Senior member
Jul 15, 2006
799
965
136
I always favoured AMD over Intel, but if this 65->105W is real, then they should fire all the marketing team and senior management who oversaw this dumpster fire. I hoped after getting better financials, they would be acting more proffessional, but this doesn't give me any hope for Strix Halo. I am hoping that will be my next APU but if they don't get their ...act together, then it will be an even better dumpster fire.

Can't manage to launch something for desktop? No biggie, go for server and launch for DT 3 months later when all software and hardware bits are in place. I mean this is a server CPU disguised in civilian clothes to fool the fanbase. There is no thought in this CPUs for DT, just for server

This really looks bad for them
 

Mahboi

Golden Member
Apr 4, 2024
1,058
1,969
96
This thread should be renamed to "Zen 5 mental gymnastics"
909lbb.jpg
 

Mahboi

Golden Member
Apr 4, 2024
1,058
1,969
96
what a pathetic launch by AMD. Seriously, learn some marketing skills AMD.
I have started believing that the marketing team at AMD knows fully how incompetent they are, and are collectively lobbying to deny anyone with competence to enter the office, lest they all get fired.
And at this point next to anyone will be more competent than them.
 

Josh128

Banned
Oct 14, 2022
1,542
2,295
106
what a pathetic launch by AMD. Seriously, learn some marketing skills AMD.

It makes you wonder what kind of little corporate "yes man" bubble they must be under, when what I assume is a million dollar plus marketing budgeted team could have been made to look stupid by some random plebs on a tech forum. Theres no doubt in my mind that many of the enthusiasts in this very thread would have advised that, at least for DIY desktop:

1.) Lying about beating 14900K and X3D in gaming is bad.
2.) Touting average IPC that in reality is greater than gen on gen perf gains due to clock regressions is bad.
3.) Leaving significant perf on the table by not utilizing previous gen TDPs is bad.
4.) Pricing these SKUs as they were with performance as it is with previous gen price/perf being so much better and 3nm competition looming is bad.

Amateurish.
 

eek2121

Diamond Member
Aug 2, 2005
3,491
5,183
136
I am absolutely shocked if that is true, but it should have been 105w from the get-go.
People overestimate the usefulness of AVX-512. The vast, vast, majority of consumers are never going to interact with it or care. Also, Intel isn't faltering that much in the

As one of the many guys that writes software that everyday folks like you use, I disagree.
 

poke01

Diamond Member
Mar 8, 2022
4,881
6,222
106
So if AMD is increasing the TDP of 9700X and 9600X with AGESA, does this mean that AGESA will also increase the 9950X performance to 1000$?

:p
Nah, they are saving that patch for Zen6 when marketing fails again.
 

Abwx

Lifer
Apr 2, 2011
12,034
4,995
136
I love how you have to keep changing the numbers and what node AMD uses in order to get the numbers to fit lmao

I provided a link that say so, so what s your point exactly ?

Doing ad hominems..?.

Numbers are right and proved by the reviews, but if you want to contradict me with actual numbers do so, i m waiting for some "explanation".

 

MS_AT

Senior member
Jul 15, 2024
941
1,866
96
With Zen 4 and Zen 5 becoming mainstream, developers have more incentive to add support for it. Especially as Intel continues to falter.
Not really. People complain that game developers require AVX2 to run newest game [Alan Wake TLOU PC port iirc] even though the instruction set is already 10 years in the market. There is a lot of software that sticks to SSE because Intel misguided politics and artificial market fragmentation.

People overestimate the usefulness of AVX-512. The vast, vast, majority of consumers are never going to interact with it or care. Also, Intel isn't faltering that much in the client market...
And we have Intel to thank for that. It's not that AVX512 is so bad, it's that Intel made it a nightmare with the fragmentation. It's like Apple would decide each M2, M2 Pro and M2 Ultra support different instruction sets... Actually Apple is in the best possible position for introducing new instruction sets, as at least through the system apps they control they can make immediate use of them.
 

poke01

Diamond Member
Mar 8, 2022
4,881
6,222
106
ouch, you guys seen that the 7950X performs better in the PS3 emulation than the 9950X? In the TPU review.
 
  • Like
Reactions: Josh128

JustViewing

Senior member
Aug 17, 2022
272
478
106
People overestimate the usefulness of AVX-512. The vast, vast, majority of consumers are never going to interact with it or care. Also, Intel isn't faltering that much in the client market...
It is completely Intel's fault. If they didn't segment it based on SKU, we might have more wider usage. AMD has proved that Intel's AVX-512 implementation is pathetic.