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

Page 95 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

A///

Diamond Member
Feb 24, 2017
4,351
3,158
136
If history is any indication, we'll continue to get random guesses and hype mongering disguised as leaks until just before release.
You're about as old as I am. Isn't that the fun part of life at this point in our lives? I'm at the edge of developing early arthritis and that's been the most exciting news in my life outside this scene. get with the flow of the world fellow old man.
 
  • Like
Reactions: Tlh97 and MadRat

Exist50

Platinum Member
Aug 18, 2016
2,452
3,102
136
You're about as old as I am. Isn't that the fun part of life at this point in our lives? I'm at the edge of developing early arthritis and that's been the most exciting news in my life outside this scene. get with the flow of the world fellow old man.
Having some non-digital hobbies helps. And for all that I complain about the endless hype trains, I'll take it over the "end of history" rhetoric from a decade ago.
 
  • Like
Reactions: moinmoin

A///

Diamond Member
Feb 24, 2017
4,351
3,158
136
Having some non-digital hobbies helps. And for all that I complain about the endless hype trains, I'll take it over the "end of history" rhetoric from a decade ago.
I do wine and old books. Prior to covid I'd go flea market hunting for unique items I knew of from decades past that most people couldn't figure out and still don't. I pray social media doesn't die out or else these morons will find a way to learn about valuable old things other sell at pittyful rates or exchange it for some modern shiny object like a soil cart. I used to do art collecting many years ago but social media grabbed onto that and made it too easy to find out info about artists and what value their work had.
 
  • Like
Reactions: lightmanek

Abwx

Lifer
Apr 2, 2011
11,557
4,349
136
Those numbers sound reasonable, if nothing else. Would be in line with other tock architectures from AMD, and 6GHz wouldn't be much of a stretch over Zen 4, especially given another couple percent from the node. 25% IPC as claimed, however, is rather more questionable. Maybe in some particular workload.

But it goes without saying, RTG has a "mixed" history to put it gently... Not sure there's any real value here beyond another dart in the board.

He didnt specify if it was ST or MT, if he s talking of ST, and of course excluding games, the improvement should be better than 15% since that s the ST IPC difference between Z2 and Z3.

If he s talking MT then better than 12-15% is enough, also excluding games, if games are included then 19% is the basis to best and in this case 25% is possible, methink that without games included in the mix 25% is quite a stretch, but who knows.

Edit : His previous statement some times ago was 19% better IPC, so we re still in the same numbers at this point.
 
Last edited:

Exist50

Platinum Member
Aug 18, 2016
2,452
3,102
136
He didnt specify if it was ST or MT, if he s talking of ST, and of course excluding games, the improvement should be better than 15% since that s the ST IPC difference between Z2 and Z3.

If he s talking MT then better than 12-15% is enough, also excluding games, if games are included then 19% is the basis to best and in this case 25% is possible, methink that without games included in the mix 25% is quite a stretch, but who knows.

Edit : His previous statement some times ago was 19% better IPC, so we re still in the same numbers at this point.
I'm not going to give him the views, but if he said specifically IPC, then that has nothing to do with MT or clocks. Zen 3 got a particularly large boost in games because of the uncore changes (more to an 8c CCX/unified L3). There've been some rumors about fabric changes with Zen 5, but nothing that I think would indicate a similar step function improvement for gaming.
 
  • Like
Reactions: Tlh97 and Thibsie

Abwx

Lifer
Apr 2, 2011
11,557
4,349
136
I'm not going to give him the views, but if he said specifically IPC, then that has nothing to do with MT or clocks. Zen 3 got a particularly large boost in games because of the uncore changes (more to an 8c CCX/unified L3). There've been some rumors about fabric changes with Zen 5, but nothing that I think would indicate a similar step function improvement for gaming.

There s ST IPC and MT IPC, that s two different things (since in MT SMT efficency has a say), at same clocks and only looking at apps Zen 3 provided 15% better perf in ST although Computerbase use only 3 benches for this test.

In MT, and same clocks, either 6C/12T or 8C/16T provide the same result, it s more mixed depending of the apps, in CB it s about 10-11%, 14% in 7ZIP and an average of 11% for 9 benches.

 
  • Like
Reactions: Thibsie

Exist50

Platinum Member
Aug 18, 2016
2,452
3,102
136
There s ST IPC and MT IPC
The only time "MT" IPC ever comes it is in the context of SMT yield. It really doesn't make sense for comparing multiple cores. And when AMD gives an IPC number on stage, it's ST 100% of the time.
In MT, and same clocks, either 6C/12T or 8C/16T provide the same result, it s more mixed depending of the apps, in CB it s about 10-11%, 14% in 7ZIP and an average of 11% for 9 benches.
That's more about the thread scaling of the particular tests than it is about IPC.
 
  • Like
Reactions: Lodix

Abwx

Lifer
Apr 2, 2011
11,557
4,349
136
The only time "MT" IPC ever comes it is in the context of SMT yield. It really doesn't make sense for comparing multiple cores. And when AMD gives an IPC number on stage, it's ST 100% of the time.

That's more about the thread scaling of the particular tests than it is about IPC.

I edited my post when you did this post and pointed SMT efficency, AMD used to have better SMT efficency than Intel, so the MT score was upped by this better behaviour, it cant be downplayed.

As for ST being what AMD states 100% of the time that s not true at all, for Zen 3 the 19% figure did include some games, wich are not single threaded, actually that was more of a MT IPC number than anything else for most apps used in the average.


26-1080.0e92feb1.png


 

Abwx

Lifer
Apr 2, 2011
11,557
4,349
136
Thanks for pointing that out, but it's more like marketing massaging the numbers than anything else. Certainly it's not how the core architects would measure things. In any case, I think you're giving these "leakers" more benefit of the doubt than they deserve. The scorn is well earned.

No, because i stated the numbers as measured by core architects, they are available at Computerbase and do not include games, also ST and MT are measured separately.

At the risk of repeating things from Z2 to Z3 apps ST IPC was improved by 15% and apps MT IPC by 11% according to Computerbase.

So the 19% once stated by RTG is within the realm of possibilities for apps, wether that s ST or MT, the most recent 25% number likely include games to boost the first number, methink that it s reasonable given that Zen 5 is supposed to be a massive overhaul of Zen 4 wich itself has a proved 12-13% better IPC in apps.
 

Exist50

Platinum Member
Aug 18, 2016
2,452
3,102
136
No, because i stated the numbers as measured by core architects
That slide is from marketing, not architects. It's just muddying the waters to boost the score a bit. Literally 64% of the datapoints from that slide are games or representative of gaming. In no way would that reflect the tracelists AMD actually uses internally (expect a lot more datacenter workloads...)

ComputerBase reported numbers from their own test suite using AMD's methodology, for the sake of comparing against those marketing claims. It's still not how that term is typically used.

Also, it's possible AMD didn't even include MT numbers at all, if those scores were generated on captured single thread traces from the games in question (possibly with a full CCX worth of L3). They don't really provide enough info to say, but it's notable that everything that specifies is tested 1T, not nT.
So the 19% once stated by RTG is within the realm of possibilities for apps, wether that s ST or MT, the most recent 25% number likely include games to boost the first number, methink that it s reasonable given that Zen 5 is supposed to be a massive overhaul of Zen 4 wich itself has a proved 12-13% better IPC in apps.
Certainly I think Zen 5 will be a bigger IPC boost then Zen 4 (though net performance might be a more interesting question). And those numbers are at least within the realm of reason. But the "source" in question has a terrible track record, and there's no indication that this time is any more believable than the past.
 
  • Like
Reactions: Lodix

Timorous

Golden Member
Oct 27, 2008
1,748
3,240
136
That slide is from marketing, not architects. It's just muddying the waters to boost the score a bit. Literally 64% of the datapoints from that slide are games or representative of gaming. In no way would that reflect the tracelists AMD actually uses internally (expect a lot more datacenter workloads...)

ComputerBase reported numbers from their own test suite using AMD's methodology, for the sake of comparing against those marketing claims. It's still not how that term is typically used.

Also, it's possible AMD didn't even include MT numbers at all, if those scores were generated on captured single thread traces from the games in question (possibly with a full CCX worth of L3). They don't really provide enough info to say, but it's notable that everything that specifies is tested 1T, not nT.

Certainly I think Zen 5 will be a bigger IPC boost then Zen 4 (though net performance might be a more interesting question). And those numbers are at least within the realm of reason. But the "source" in question has a terrible track record, and there's no indication that this time is any more believable than the past.

Ian tested SPECrate in the Anandtech Zen 3 review and got 19% single thread uplift and 10% multi thread uplift. For real world they got a 24% uplift in their suite.
 

Exist50

Platinum Member
Aug 18, 2016
2,452
3,102
136
Ian tested SPECrate in the Anandtech Zen 3 review and got 19% single thread uplift and 10% multi thread uplift. For real world they got a 24% uplift in their suite.
Nonetheless, if AMD used the same suite as above, but less games, the number would be lower. They could have just quoted SPEC if they wanted, so I'm not sure why the hoopla. Certainly they're more likely to use SPEC day to day than 2/3 gaming traces.
 
  • Like
Reactions: Lodix

Timorous

Golden Member
Oct 27, 2008
1,748
3,240
136
Nonetheless, if AMD used the same suite as above, but less games, the number would be lower. They could have just quoted SPEC if they wanted, so I'm not sure why the hoopla. Certainly they're more likely to use SPEC day to day than 2/3 gaming traces.

Because it was a marketing slide and it captures a lot of real world use cases which is far more useful for selling a consumer product than quoting SPEC benchmarks that a lot of people have not even heard of.
 

Exist50

Platinum Member
Aug 18, 2016
2,452
3,102
136
Because it was a marketing slide and it captures a lot of real world use cases which is far more useful for selling a consumer product than quoting SPEC benchmarks that a lot of people have not even heard of.
Well that begs the question then. To what degree does SPEC represent real world use cases today? But yes, my point was that it is marketing, and should not be taken as representative of AMD's internal measurements, or how they measure/use the term "IPC".
 
  • Like
Reactions: Lodix

Abwx

Lifer
Apr 2, 2011
11,557
4,349
136
ComputerBase reported numbers from their own test suite using AMD's methodology, for the sake of comparing against those marketing claims. It's still not how that term is typically used.
You are lying here, dunno why you re acting like this.

Computerbase use two versions of CB and Pov RAY for ST IPC.

For MT IPC they use 7 ZIP, AgitSoft, Blender, CB R15, CB R20, Corona 1.3, Digicortex, Handbrake and PovRAY.

As you can see there s no games here and ST and MT scores are not mixed, there s separate averages.

Obviously you didnt even look at the link i posted and are completely wrong, but nevermind...

 

Abwx

Lifer
Apr 2, 2011
11,557
4,349
136
How does Zen 3 IPC gain have anything to do with Zen 5 uarch? Can you discuss it somewhere else or do we have this thread blocked again?

It went there because RGT stated that Zen 5 would provide better IPC improvement comparatively to Zen 4 than what was brought by Zen 3 comparatively to Zen 2.

So i provided Zen 3 numbers to have an idea of what is supposed to be bested, but as usual there s some people with few knowledge on the matter dismissing those numbers for whatever reason, likely that it s some kind of trolling, hence the lengthy debate...
 
  • Like
Reactions: Tlh97 and Mopetar

eek2121

Diamond Member
Aug 2, 2005
3,100
4,398
136
They didn't, it's just physdes flexing.

How many times do I have to tell everyone that AMD does a proper tock every other gen (they're odd-numbered) and every even is iterative.

Regarding tick-tock, This is simply not true.

Zen1, Zen+, Zen 2, and all variants = family 17h

Zen 3, Zen3+, Zen 4, and all variants are 19h

Zen 5 is 1Ah.

Of these, Zen 3 was considered a ‘tock’ due to a block level rebuild of the silicon which likely changed enough in terms of errata and performance characteristics it could be considered breaking.

Zen 5 supposedly is the next rebuild.

Oh and while the core in Zen 4 has evolved, it is still a direct descendant of the core in Zen 1, sharing many similarities, limitations, pitfalls, etc. It remains to be seen if this changes with Zen 5 or if it truly is a new design.
 

Exist50

Platinum Member
Aug 18, 2016
2,452
3,102
136
You are lying here, dunno why you re acting like this.
When I said, "same methodology", I mean humoring the idea that "MT IPC" is an actual metric that can be measured in the way AMD marketing presented. They just used a different set of tests for the sake of comparison. That's not some standard part of their reviews, is the point.

Anyway, going back to the original purpose of assessing rumor mongers, there's any number of things from their history you can point to to know they're making it all up. They take advantage of you giving them this benefit of the doubt for views (money), even if they have to lie to your face to do so. It's just not worth anyone's time to take such claims seriously.
 

Abwx

Lifer
Apr 2, 2011
11,557
4,349
136
When I said, "same methodology", I mean humoring the idea that "MT IPC" is an actual metric that can be measured in the way AMD marketing presented. They just used a different set of tests for the sake of comparison. That's not some standard part of their reviews, is the point.

I thought that you noticed that Computerbase metrics ,separating ST and MT numbers as well as not using games, are completely different from AMD s (FTR they did an average for games starting with Zen 4) but since you didnt notice there s nothing i can add on the subject.

Anyway, going back to the original purpose of assessing rumor mongers, there's any number of things from their history you can point to to know they're making it all up. They take advantage of you giving them this benefit of the doubt for views (money), even if they have to lie to your face to do so. It's just not worth anyone's time to take such claims seriously.

If we look at Zen 4 improvements most significant contributions in order of importance come from the front end, although it was unchanged globally, LSU and then branch prédiction, exe engine and cache improvements are the two least meaningfull in the list of contributions.

Now if we look at Zen 5 stated improvements AMD pointed a re-pipelined front end and a wider issue, the latter requiring forcibly a better branch prédiction as well as an improved LSU to cope with the augmented scheduling of instructions.

So they are reworking parts that are instrumental to extract more IPC, wich tell us that a significantly bigger improvement than what was brought by Zen 4 is more than likely.
 
Last edited:

Exist50

Platinum Member
Aug 18, 2016
2,452
3,102
136
I thought that you noticed that Computerbase metrics ,separating ST and MT numbers as well as not using games, are completely different from AMD
Again, copying the methodology, but different test suite, hence different results.
So they are reworking parts that are instrumental to extract more IPC, wich tell us that a significantly bigger improvement than what was brought by Zen 4 is more than likely.
Yes, Zen 5 in all likelihood will be a bigger IPC jump than Zen 4. Overall performance, however, might not be, if they don't get a meaningful frequency boost as some are alleging. Would fit better into the pattern of Zen 3, which makes sense given that that was also a new architecture, though by a different team, iirc. Clearly they originally wanted to use N3.
 
  • Like
Reactions: Tlh97 and A///

Abwx

Lifer
Apr 2, 2011
11,557
4,349
136
Again, copying the methodology, but different test suite, hence different results.

Rather than aknowledging that you were wrong you keep being in denial, different methodologies hence different results, theres s no separation of ST, MT and games in AMD s slide, while Computerbase separate ST, MT and games and provide averages for each of these cases, there s really no worst blind that the one that keep his eyes closed purpotedly...

Yes, Zen 5 in all likelihood will be a bigger IPC jump than Zen 4. Overall performance, however, might not be, if they don't get a meaningful frequency boost as some are alleging. Would fit better into the pattern of Zen 3, which makes sense given that that was also a new architecture, though by a different team, iirc. Clearly they originally wanted to use N3.

There will be a significant uplift because they ll use 4nm instead of 5nm, wich will give a lttle room to improve perf, to wich will be added some significant IPC uplift, they know well for ages that they couldnt rely on a vastly better process for this gen, hence they logically did put much more efforts in IPC.

I guess that you re not trying to have a logical reasoning but rather are hoping that the progress will be meaningfull to better help intel quiting the rear view mirror, unfortunaly that likely wont happen because Zen 5 look like a big push architecturally wise, bigger than all previous iterations of Zen.
 
  • Like
Reactions: Tlh97

Exist50

Platinum Member
Aug 18, 2016
2,452
3,102
136
Rather than aknowledging that you were wrong you keep being in denial, different methodologies hence different results, theres s no separation of ST, MT and games in AMD s slide
You should reread my comments, because I addressed all of that several times now. I'll ignore the rest of your attempted provocation.
There will be a significant uplift because they ll use 4nm instead of 5nm, wich will give a lttle room to improve perf
4nm does not give a significant improvement vs 5nm, and that's even assuming all these design/architecture changes have no impact on the critical path, which is unlikely.
they know well for ages that they couldnt rely on a vastly better process for this gen
When originally planning for the core, they most likely assumed a healthy N3 with noticeable improvements over 5nm. But TSMC slipped, and so they had to retarget N4 instead.
 

Abwx

Lifer
Apr 2, 2011
11,557
4,349
136
You should reread my comments, because I addressed all of that several times now. I'll ignore the rest of your attempted provocation.

Anything that prove you wrong is indeed provocation, i produced hard numbers while, as usual, you are left manipulating numberless and hollow sentences.

4nm does not give a significant improvement vs 5nm, and that's even assuming all these design/architecture changes have no impact on the critical path, which is unlikely.

4nm doesnt need to bring much improvement, just enough to keep clocks at say 95% and get the same power given that this latter will be increased due to bigger uarch and higher throughput/Hz, 0.95X the frequency allow for 12-13% lower power, to wich will be added the small improvement over 5nm.

This will end at 0.8x the power and would allow same power at at better than 1.25x higher throughput assuming a theorical 25% better IPC in MT.

When originally planning for the core, they most likely assumed a healthy N3 with noticeable improvements over 5nm. But TSMC slipped, and so they had to retarget N4 instead.

There s different flavours of N4, some that have efficency close enough to N3, FI N4P has 6% better perf at isowatt than N4, this allow to reduce power by about 15% at isofrequency comparatively to N4.

Relatively to N5, wich interest us most, it allow 11% better perf/isowatt or 24% lower power at isoperf, there s enough room for Zen 5 to have 31% better perf at same TDP than Zen 4.

Guess that you did a mental build up to find all possiblities that could make Zen 5 being bottlenecked by whatever suits your hopes, but those numbers just crumble your house of cards theories...
 

Exist50

Platinum Member
Aug 18, 2016
2,452
3,102
136
N3, FI N4P has 6% better perf at isowatt than N4, this allow to reduce power by about 15% at isofrequency comparatively to N4.
To just point out the most glaring of many inaccuracies here, 6% better perf/watt means ~6% lower power at iso-frequency, not 15%. Fabs would call the other metric performance. But if I have to once again repeat common knowledge, I think this conversation no longer has value.