Discussion Speculation: Zen 4 (EPYC 4 "Genoa", Ryzen 7000, etc.)

Page 43 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Vattila

Vattila

Senior member
Oct 22, 2004
799
1,351
136
Except for the details about the improvements in the microarchitecture, we now know pretty well what to expect with Zen 3.

The leaked presentation by AMD Senior Manager Martin Hilgeman shows that EPYC 3 "Milan" will, as promised and expected, reuse the current platform (SP3), and the system architecture and packaging looks to be the same, with the same 9-die chiplet design and the same maximum core and thread-count (no SMT-4, contrary to rumour). The biggest change revealed so far is the enlargement of the compute complex from 4 cores to 8 cores, all sharing a larger L3 cache ("32+ MB", likely to double to 64 MB, I think).

Hilgeman's slides did also show that EPYC 4 "Genoa" is in the definition phase (or was at the time of the presentation in September, at least), and will come with a new platform (SP5), with new memory support (likely DDR5).

Untitled2.png


What else do you think we will see with Zen 4? PCI-Express 5 support? Increased core-count? 4-way SMT? New packaging (interposer, 2.5D, 3D)? Integrated memory on package (HBM)?

Vote in the poll and share your thoughts! :)
 
Last edited:
  • Like
Reactions: richardllewis_01
Sort by date Sort by votes
DisEnchantment

DisEnchantment

Golden Member
Mar 3, 2017
1,601
5,779
136
https://www.reddit.com/r/Amd/comments/myj91t
Click to expand...
I can't help but wonder if there is some truth to this report from RGT. Although I don't really like this kind of reporting, Paul has been right many times so I am willing to consider this could be a possibility.
There are no events listed in AMD's IR calendar, total silence til mid of the year. So we wont be finding out any time soon.
I agree on the reasoning as well, Zen 4 on 5nm would use the new Fab 18 which does not share anything with other fabs using N6/N7. N6 and N7 share equipment so I guess capcaity remains an issue if they use N6.
 
  • Like
Reactions: Tlh97, lightmanek, uzzi38 and 2 others
Upvote 0 Downvote
M

Mopetar

Diamond Member
Jan 31, 2011
7,831
5,979
136
Paul reports on a lot of rumors so he's right a lot only because he reports on the rumors that pan out as much as the ones that don't.

If Zen 4 is well along and AMD suspects that Intel is going to have a good showing with their future products there's little reason to devote time to a stop-gap product.

The other side of this could be that AMD was expecting more from Intel in their most recent release and had prepared Zen 3+ as a potential response to regain supremacy. That no longer seems necessary as AMD's biggest issue is being able to supply demand and Zen 3+ does nothing to help that.
 
Upvote 0 Downvote
S

scineram

Senior member
Nov 1, 2020
361
283
106
Indeed the reasoning does make sense. But only if it’s for AM4 still, not launching AM5. Which would be the disappointment scenario for me anyway.
 
Upvote -1 Downvote
K

Kedas

Senior member
Dec 6, 2018
355
339
136
If Alder Lake would only be available on their new DDR5 platform then there is no point for AMD to release an Zen3+ on DDR4 to compete with it.
AMD would still have the best DDR4 platform.

And then we don't even talk about the fact that AMD is selling everything they can produce. So what would change after Zen3+ release, nothing.
 
  • Like
Reactions: Tlh97, Makaveli, ryan20fun and 2 others
Upvote 0 Downvote
H

HurleyBird

Platinum Member
Apr 22, 2003
2,684
1,267
136
Kedas said:
If Alder Lake would only be available on their new DDR5 platform then there is no point for AMD to release an Zen3+ on DDR4 to compete with it.
AMD would still have the best DDR4 platform.
Click to expand...

If Zen3+ beats Alder Lake, but Zen3 doesn't (or if it moves the needle from questionable to not), then I disagree.

There's something significant about a last-gen platform beating the next-gen platform of your competitor. And AM4/DDR4 will be around in the market for quite a while. Making the final product line on that platform perform as well as possible would hardly be irrelevant.

Which isn't to say that focusing on Zen4 isn't the right move even if there are reasons for Zen3+. If AMD can speed up Zen 4 development by say, three months, then that sounds like a good trade-off. There's nothing stopping AMD from restarting work on Zen3+ after Zen 4 launches either, if it seems like a good market move and there's already a good chunk of sunk cost.
 
Upvote 0 Downvote
M

mikk

Diamond Member
May 15, 2012
4,133
2,136
136
I'm expecting no more than a slighty higher clocked Zen 3 refresh. Maybe they have realized that using a new generation naming for a slightly higher clocking Zen 3 is not a good idea.
 
  • Like
Reactions: Mopetar
Upvote 0 Downvote
Ajay

Ajay

Lifer
Jan 8, 2001
15,428
7,847
136
Tarkin77 said:
https://twitter.com/x/status/1386846543873933312

sounds interesting....

"In addition, the new CPU in the second half of this year will be very interesting."

"Yes, very powerful. Especially in games."

would could that be ??
Click to expand...
Uh, this guy has been sure of Warhol as early as April 10th:
https://twitter.com/x/status/1380767062038052865

Warhol makes sense to me, unless Zen4 is due out sooner than we expect. An important point is that this decision would have been made quite some time ago. AMD isn't making decisions on whether or not to design and make masks for a given chip on how well a particular Intel CPU is performing. They are making decisions based on their own internal time frames; perhaps with some regard to Intel's roadmap (which they know better than we do).
 
  • Like
Reactions: Tlh97 and Elfear
Upvote 0 Downvote
DisEnchantment

DisEnchantment

Golden Member
Mar 3, 2017
1,601
5,779
136
Thibsie said:
IMO not a big.little as we know it.
Patents show (if those are indeed used) a quite smart way of implementing this.
Click to expand...


Yeah I did write about this patent somewhere back

DisEnchantment said:
A more consequential patent is this actually, which I also posted some pages back

US Patent has been awarded was filed on Oct 27 2017

But what is interesting is that AMD applied for another similar patent again. It seems likely that they found additional things to do the with it and applied for it again. Filed June 25 2020.
Not yet awarded, still in application state but will surely be awarded as it is just a continuation of awarded patent.

Similar abstract but 20 additional claims.

You can read both, they are awesome.

View attachment 37823View attachment 37824



Like I wrote in my earlier post, there are shared registers between the big and small cores.
In low power mode the small cores are running only a small subset of low powered instructions, when a complex instruction is encountered a trap occurs and the big core takes over seamlessly.
OS is not even aware that all of this is happening. It is basically the same core to the OS.
Because of shared Registers L2 etc, the number of small cores is exactly the same as the number of big cores and at any time either the big or the small core is active not both.
In my opinion this could be done fairly cheaply in terms of die real estate, they could power gate selected blocks in the current core and add special execution ports, L/S ports and other tidbits and thats all.
I imagine in low power the first target would be to power gate the complex decoder block, leaving just simple decoding blocks, power gate all the executions ports except the most simple and power efficient one, power gate a chunk of the register file, and so on.
More innovative than the big.LITTLE imo.

I suppose this opens up new possibilities on what the CPU can do.
It can execute multiple ISA with the same core :blush: . illegal opcode trap in one core changes the CPU to another and continue execution. And OS is not aware.
Dreams...
Click to expand...

I do find their solution to be unique and if there are several follow up patents it usually end up happening.
 
  • Like
Reactions: Gideon, Tlh97, Thibsie and 3 others
Upvote 0 Downvote
Ajay

Ajay

Lifer
Jan 8, 2001
15,428
7,847
136
mikk said:
https://twitter.com/x/status/1387007700895031299

Bullsh1t_Buster also says Warhol has been cancelled.
Click to expand...

+ 100-200 MHz will be kind of worthless 5000 XTs. So, it's really just a marketing exercise at this point if true. Also, this means Warhol was cancelled something like 2 years ago.

Also, boo hoo, was hoping for a good AM4 replacement for my 3900X. Guess I'll be waiting 2-3 years since I'll need a new board and ram as well as a new CPU.
 
Upvote 0 Downvote
moinmoin

moinmoin

Diamond Member
Jun 1, 2017
4,944
7,656
136
We simply know nothing. Chances are all the CPU rumors, if founded on something real anyway, are from the time when AMD was still more leaky, so well before the launch of the whole Zen 3 line. The last actual data we got from leaks were the specs for RDNA2 chips due to driver and firmware data from several sources. Even that has dried up since.
 
Upvote 0 Downvote
C

CakeMonster

Golden Member
Nov 22, 2012
1,389
496
136
HurleyBird said:
Which isn't to say that focusing on Zen4 isn't the right move even if there are reasons for Zen3+. If AMD can speed up Zen 4 development by say, three months, then that sounds like a good trade-off. There's nothing stopping AMD from restarting work on Zen3+ after Zen 4 launches either, if it seems like a good market move and there's already a good chunk of sunk cost.
Click to expand...

I would consider it very unlikely that it can be sped up, I'd rather assume that they push the release back to build up inventory. I'd love to be wrong though.

I guess we've been conditioned to not expect yearly releases, which is why I never really believed in "Zen3+". However, if Zen4 is late '22 or pushed into '23 then I would expect something IF Intel is also late. We don't really have the info to assess the DDR5 or Alder Lake situation so we're really in the dark here. I feel less sure about making guesses than in a long time. I guess that's why people are hanging on to people posting fanfic on Twitter and YT too.
 
Upvote 0 Downvote
uzzi38

uzzi38

Platinum Member
Oct 16, 2019
2,620
5,870
146
mikk said:
https://twitter.com/x/status/1387007700895031299
https://twitter.com/x/status/1387008366338220034

Bullsh1t_Buster also says Warhol has been cancelled. Interestingly he says AMD is working on a big+little generation. I think it's the future, Intel is just the first with Alder Lake.
Click to expand...
This article is written by the owner of that Twitter account: https://moepc.net/amd-zen-5-codenam...ig-little-architecture-using-the-3nm-process/

Few extra details that weren't in the Tweets.
 
  • Like
Reactions: Elfear, Tlh97 and Mopetar
Upvote 0 Downvote
E

eek2121

Platinum Member
Aug 2, 2005
2,930
4,025
136
Ajay said:
+ 100-200 MHz will be kind of worthless 5000 XTs. So, it's really just a marketing exercise at this point if true. Also, this means Warhol was cancelled something like 2 years ago.

Also, boo hoo, was hoping for a good AM4 replacement for my 3900X. Guess I'll be waiting 2-3 years since I'll need a new board and ram as well as a new CPU.
Click to expand...

For what it is worth, AMD appears to have Warhol as a stand-in for possible 5nm capacity issues OR delayed DDR5 availability. DDR5 looks to be “on track”. Warhol was said to be a DDR4 product and contrary to popular belief, the new die and the old die would not easily interchangeable.

The cancelation rumors may also be completely wrong.

For what it is worth, Only 2 folks have said that it is canceled thus far, and at least one of them is a clown.l and could have gotten it from the other or vice versa.
 
  • Like
Reactions: Thibsie, lightmanek and Tlh97
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

COMPANY
RESOURCES
FOLLOW
Top Bottom