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

[Vattila]

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).

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!

 

[Markfw]

Hassan said he'd reserve a separate article for now. What's bs about his infos btw?

He is not a writer for that website is he ? I do not know the person, but anything in a comments section, especially that section sure looks like trash. I can't even use the correct words to describe it.

 

[Vope45]

He is not a writer for that website is he ? I do not know the person, but anything in a comments section, especially that section sure looks like trash. I can't even use the correct words to describe it.

He is the writer

 

[Timorous]

According to wccf's Hassan motherboard maker source, Zen 4 is still behind AL and RL in cpu sensitive games: https://wccftech.com/amd-ryzen-7000-desktop-cpus-x670-motherboards-launch-on-15th-september/


Look at the comments section: RL > Zen 4 in both latency and bw running the same DDR5 JEDEC speeds . 3 games were tested SOTR, FH5 and FFXIV plus FSU.

That would make it slower than the 5800X3D, especially in FFXIV.

 

[Markfw]

He is the writer

Well, isn't that site not highly regarded here ? and a few of the comments after your post support the fact he is wrong.

 

[Asterox]

According to wccf's Hassan motherboard maker source, Zen 4 is still behind AL and RL in cpu sensitive games: https://wccftech.com/amd-ryzen-7000-desktop-cpus-x670-motherboards-launch-on-15th-september/


Look at the comments section: RL > Zen 4 in both latency and bw running the same DDR5 JEDEC speeds . 3 games were tested SOTR, FH5 and FFXIV plus FSU.

There is no comments, even with UBlock Origin disabled i dont see(Chroome of Firefox same story) any comments.With adblock disabled, that website is classic garbage+a bunch of ads.

 

[Carfax83]

According to wccf's Hassan motherboard maker source, Zen 4 is still behind AL and RL in cpu sensitive games: https://wccftech.com/amd-ryzen-7000-desktop-cpus-x670-motherboards-launch-on-15th-september/


Look at the comments section: RL > Zen 4 in both latency and bw running the same DDR5 JEDEC speeds . 3 games were tested SOTR, FH5 and FFXIV plus FSU.

I'm honestly not surprised. Alder Lake was straight up brutal against Zen 3 (even with enhanced cache) in a few CPU intensive titles like CBP 2077 and the Riftbreaker, which demonstrated that Alder Lake is just on a whole other level.

Also, Intel's IMC is closer to the CPU cores than Zen 4s will be due to a monolithic design.

 

[Vope45]

There is no comments, even with UBlock Origin disabled i dont see(Chroome of Firefox same story) any comments.With adblock disabled, that website is classic garbage+a bunch of ads.

Hassan said the source who gave him zen 4 launch and announcement dates also sent in performance number and overall impression.

 

[Markfw]

Hassan said the source who gave him zen 4 launch and announcement dates also sent in performance number and overall impression.

Based on what I have seen, what he said is totally wrong.

 

[Vope45]

Based on what I have seen, what he said is totally wrong.

What have you seen?

 

[Markfw]

What have you seen?

Look, read this thread, all the replies, and all the other "leaks". I am not going to debate this anymore. Wait for real reviews, and lets see if I am right,. Post then.

 

[Vope45]

Look, read this thread, all the replies, and all the other "leaks". I am not going to debate this anymore. Wait for real reviews, and lets see if I am right,. Post then.

I sure hope zen 4 will be on par or better than RL it's just the gap is wide.

 

[Saylick]

To save everyone some time so that they don't have to dig through the dumpster fire which is the WCCFTech comment section (seriously, I hate that website), here's the pertinent comments from Hassan. Looks like his source is the mobo manufacturers (no surprise, considering it sure as heck won't be someone inside AMD).

 

[Markfw]

To save everyone some time so that they don't have to dig through the dumpster fire which is the WCCFTech comment section (seriously, I hate that website), here's the pertinent comments from Hassan. Looks like his source is the mobo manufacturers (no surprise, considering it sure as heck won't be someone inside AMD).

View attachment 65332
View attachment 65333
View attachment 65334
View attachment 65335
View attachment 65336

Thanks for that ! I still don't believe it, I will wait for real benchmarks.

And I have 32 gig of DDR5 6600 cl34 coming Friday to be prepared.....

 

[RnR_au]

And I have 32 gig of DDR5 6600 cl34 coming Friday to be prepared.....

Yeah... good move. There could be a shortage of various components on AMD's release day or shortly after.

 

[coercitiv]

here's the pertinent comments from Hassan. Looks like his source is the mobo manufacturers (no surprise, considering it sure as heck won't be someone inside AMD).

Rhetorical question: if he's convinced about the source and the information, why is the leak buried in the comments section instead of generating income and fame on the homepage?

Even the framing is weak, imagine this title on the homepage:
Zen4 behind Zen3D in gaming, according to industry sources

 

[Saylick]

Rhetorical question: if he's convinced about the source and the information, why is the leak buried in the comments section instead of generating income and fame on the homepage?

Even the framing is weak, imagine this title on the homepage:
Zen4 behind Zen3D in gaming, according to industry sources

Honestly? I think it's because he's trying to protect his motherboard manufacturers, e.g. there's been a few times he's pulled articles at the request of his sources. For what it's worth, WCCFTech used to never get review samples from AIBs and the like because they weren't seen as a tech website to be taken seriously (I still don't think they should be), but I guess over time they've made relationships with various manufacturers who gave them a shot due to the amount of traffic they garner. Those manufacturers must have taken them more seriously over time so I'm guessing those relationships became something more legitimate judging by how much I see Hassan retweet a bunch of marketing tweets from ASUS, MSI, Gigabyte, etc. I think he has something real to lose if he sours those relationships, i.e. those review samples. No review samples for WCCFTech means their credibility as a real tech website goes down (not that they had much to begin with), but they are trying their hardest to build up whatever credibility they have.

It wasn't long ago where WCCFTech exclusives were literally whatever their authors pulled out of their arse, but they have slowly stopped doing that due to aforementioned relationships and largely because there's plenty of juicy clickbait-able content that Twitter leakers post these days such that they don't have to make it up themselves.

 

[tamz_msc]

Here's HWUB's comparison of the 5800X3D and 12900K at 1440p:



Based on this, I believe what Hassan is saying. It is very likely that Zen 4 will basically match the 5800X3D.

 

[coercitiv]

Honestly? I think it's because he's trying to protect his motherboard manufacturers, e.g. there's been a few times he's pulled articles at the request of his sources. [...]

Fair enough, let's see if his behavior stays consistent with this motivation.

 

[exquisitechar]

I wouldn’t find it surprising at all if Raphael was on par with or slightly slower than ADL-S in gaming, but we’ll see.

 

[Timorous]

Here's HWUB's comparison of the 5800X3D and 12900K at 1440p:

View attachment 65342

Based on this, I believe what Hassan is saying. It is very likely that Zen 4 will basically match the 5800X3D.

Perhaps but the performance profile will be different. I doubt Zen 4 will catch up in titles like ACC, MSFS, Path of Exile and so on. The titles where the X3D has 30+% performance gains over Zen 3 but in the titles that barely move the needle or that only see a 5% or so boost with the X3D I expect Zen 4 to be faster. If that nets out to rough parity across a wide range of games I could believe that although I think on an average like HUB do Zen 4 will be a few % faster.

 

[Hans de Vries]

Zen 4: Cores, boost frequency L2/L3 cache.
So with PBO an extra 150 MHz to 5.85 GHz?

https://twitter.com/x/status/1555082501630361600

 

[Zucker2k]

I always wanted to believe those who saw Raphael-X as a backup product to counter RPL-S in the desktop space but with the speed at which AMD plans to release, right after the release of Raphael, it makes one wonder. If I'm not mistaken, it'll be the first time AMD releases chips in such short interval. The question is: Why?

My answer: Well, what are 3D caches well known for in the desktop space?

So, it seems believable to me that vanilla Raphael would be overshadowed in the desktop gaming space by both ADL-S and RPL-S and therefore needing Raphael-X to mitigate any damage in the hotly contested gaming arena. This will jive with what Hassan is reporting.

On the other hand, if there's no chip releases from either company till 2024, then there's no need to sit on a Raphael-X release after RPL-S drops. The only key question here will be: Why release Raphael-X so soon if vanilla Raphael can take on RPL-S, especially, on its own?

 

[inf64]

Perhaps but the performance profile will be different. I doubt Zen 4 will catch up in titles like ACC, MSFS, Path of Exile and so on. The titles where the X3D has 30+% performance gains over Zen 3 but in the titles that barely move the needle or that only see a 5% or so boost with the X3D I expect Zen 4 to be faster. If that nets out to rough parity across a wide range of games I could believe that although I think on an average like HUB do Zen 4 will be a few % faster.

On top of that, AMD will have Vcache Zen4 gaming parts as well. These should outperform 5800X3D by at least 15% and put Zen4 with Vcache firmly on top of the gaming performance.

 

[Karnak]

Based on this, I believe what Hassan is saying. It is very likely that Zen 4 will basically match the 5800X3D.

The 5800X3D is on average like 15% faster than the 5800X. With +8-10% IPC (very likely more in gaming due to double L2) and like +15% clocks Zen4 will easily outperform the 5800X3D.

Not in all titles ofc since there are games where the 5800X3D is >30% faster but on average? Absolutely.

 

[FangBLade]

The 5800X3D is on average like 15% faster than the 5800X. With +8-10% IPC (very likely more in gaming due to double L2) and like +15% Zen4 will easily outperform the 5800X3D.

Not in all titles ofc since there a games where the 5800X3D is >30% faster but on average? Absolutely.

And l2 cache was already tested in AIDA and it has excellent latency, combined with already excellent 3dcache + higher clocks, Zen 4-3d should have gaming crown easily.