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!

 

[CakeMonster]

I said a while ago in this thread, that if AMD doesn't raise cores from 16 to 24 they could lose the multithreaded war to the competition. In hindsight, my observation was quite myopic because it failed to take into account the one other way AMD could attain parity or superiority

In hindsight (I wanted more cores too, while not realizing I probably wouldn't need it yet), its not that surprising that AMD didn't want to deal (yet) with the challenges of adding a 3rd CCX nor design complications of more total threads/cores without losing performance in overhead.

 

[CakeMonster]

Greymon55 on Twitter is hinting at which models get V-cache... Looks like it's the 7800X, 7900X, and 7950X.

If true.... *deep inhale*

Spoiler

View attachment 66876

https://twitter.com/x/status/1564807986497081346

https://twitter.com/x/status/1564808672282558465

I want to believe, but if its at least 6 months out, then how can we count on this not changing even if its not made up? I mean, remember when 3080 released and the rumor mill was full of upcoming models with 12GB, 20GB, and Ti? They didn't arrive weeks after, they arrived 12 months or longer after... and people who delayed buying based on rumors really screwed themselves considering the prices and availability situation.

 

[Yosar]

Always amazes me how Intel squad think they can have their DDR4/DDR5 cake and eat it too when talking about pure value builds but also pure performance builds.

Ohh... it's a long history for intel squad. Like taking performance of 'K' models but price of 'F' and 'non K' models. Like there was no difference in specification between these models (max boost, max all core boost, power limits, iGP etc).

 

[Timmah!]

Greymon55 on Twitter is hinting at which models get V-cache... Looks like it's the 7800X, 7900X, and 7950X.

If true.... *deep inhale*

Spoiler

View attachment 66876

https://twitter.com/x/status/1564807986497081346

https://twitter.com/x/status/1564808672282558465

Awww, this sucks, if true. CES is usually early January, right?

I presume if its revealed at that date, then another month till the start of the sale, plus the availability might be bit scarce right away, so we might be looking at February, March.

That creates conundrum - i would probably want to upgrade to Zen5 down the road, since i think it will increase core count. Zen 5 will supposedly come sometimes in 2024, say 2 years from now, maybe even sooner. So in a way, Zen 4 might be sort of stop-gap for me (since i usually dont tend to change CPUs each generation).

Does it make sense under such condition to wait another half year for v-cache chip, when i can have vanilla chip a month from now, ultimately use it for half year longer and even pay less money for it (presuming 3D will be more expensive)?

Decisions, decisions

 

[jpiniero]

Awww, this sucks, if true. CES is usually early January, right?

I thought they were usually later but it starts Jan 5th next year.

 

[Yosar]

Please do you know the settings that Ram run in the ADL system? Do you know if it ran at CL30? What about the other timings? It's not far fetched for any enthusiast worth their salt to understand that putting an AMD EXPO tuned RAM into a non EXPO rated bios, much more a non AMD system, isn't going to run optimally. This is not so hard to understand. The 7600x run optimally. The ADL didn't. Add the cherry-picked games on top and you should understand why your "winning in games" comment is not only laughable but smacks of gullibility since these are not even third-party reviews but AMD slides. If nothing at all, history should teach us to exercise some patience until we get reviews from other sources since this isn't the first time AMD has pulled this sort of stunt with their gaming numbers.

Man, you're definition of grasping a straw.
If anand will continue their policy in test you won't see Alder Lake tested with 6000 DDR5. They test (and rightfully) with rated speed by chip producer, so 4800 for Alder Lake and 5200 for Zen 4 (and with 5600 for RL). It will be even worse for AL. The same will be in many other reviews, because that's DDR5 RAM speed both producers guarantee at stock.
Overclocking RAM is pita (especially testing) and not many people do it. They wanna know what performance is when they put their kit in mainboard go to BIOS set profile and that's all. That's why reviews (and again rightfully) don't optimize usually any RAM settings.

And latency is not good indicator of anything. It's only for the first block of data from memory, and you have to transfer many many blocks at one time. That's one of the reasons why Rocket Lake was no match for Zen 3 in games, despite having 10 ns better latency than Zen 3.

 

[DisEnchantment]

As you said, that slide can't be directly applied to Dragon Ridge or Phoenix Point.
What we know based on the AMD slide and Det0x's data is this:

	5950x	7950x (AMD slide)	7700x (calculated)
TDP(PPT)	65W (88W)	65W (88W)	~35W (44W)
Frequency	2377-2402 MHz	3803-3843 MHz	3803-3843 MHz
CB R23	16,210	28,205	~14,103

R9 6900HS tested at ComputerBase (TDP gemessen -> TCP measured).
View attachment 66882
So my expectation is that 8C16T Zen4 will manage comparable score at 1/2 power, but absolute performance at the same TDP will increase by ~20-30%.
If you want true power, then you have to buy a mobile 16C32T Dragon Ridge.

Here some data from me for GB 5.4.5 on Linux with 65W(88W) ECO mode (I always have this on always)

Gigabyte Technology Co., Ltd. X570 AORUS PRO - Geekbench

Benchmark results for a Gigabyte Technology Co., Ltd. X570 AORUS PRO with an AMD Ryzen 9 5950X processor.

browser.geekbench.com

Above is all core load.
If we get 50% perf at same power that is insane or <50% lesser power with same score.
Dev Laptops with Dragon Range are just going to be insane. Better than a stock 5950X.

 

[therealmongo]

Btw, limiting access to your profile wasn't a great idea because I was still able to find this "gem" of yours:

Frankly, Alder Lake is so far ahead of Zen 3 that Zen 4 is unlikely to do anything more than catch up to that 2021 chip.

Shady by name shady by nature

 

[FangBLade]

Shady by name shady by nature

his predictions are that AMD will return to the bulldozer era, I can't wait to see how it all plays out in the future.

 

[therealmongo]

My guess is that they had pretty good intel (ahem) about the 13900K and wanted to make sure they didn't end up a few % below it (see the unconfirmed GB scores that were extremely close).

No worries, KFC is on its way, finger licking good



But come on peeps

what I want to know,

what I really really really want to know.

Is what are userbenchmark going to do

 

[Zucker2k]

Awww, this sucks, if true. CES is usually early January, right?

I presume if its revealed at that date, then another month till the start of the sale, plus the availability might be bit scarce right away, so we might be looking at February, March.

That creates conundrum - i would probably want to upgrade to Zen5 down the road, since i think it will increase core count. Zen 5 will supposedly come sometimes in 2024, say 2 years from now, maybe even sooner. So in a way, Zen 4 might be sort of stop-gap for me (since i usually dont tend to change CPUs each generation).

Does it make sense under such condition to wait another half year for v-cache chip, when i can have vanilla chip a month from now, ultimately use it for half year longer and even pay less money for it (presuming 3D will be more expensive)?

Decisions, decisions

Go for vanilla Zen 4 on day one, if you can get it, then you can sell it down the line to offset the cost of an X3D when those are released. I'm not the one making the purchase but I'm feeling the headache from just reading your posts.

 

[Kocicak]

If we ignore all that high frequency copper melting inferno happening now and compare the single core output at nice cool enviromentally friendly 3600 MHz, we get this:



If I remember correctly, AMD stated that the new CPUs are 9% better at Cinebench than 5000 series CPUs. That would mean 508 points for Zen 4 - that is still slower than Alder lake.

 

[In2Photos]

If we ignore all that high frequency copper melting inferno happening now and compare the single core output at nice cool enviromentally friendly 3600 MHz, we get this:

View attachment 66900

If I remember correctly, AMD stated that the new CPUs are 9% better at Cinebench than 5000 series CPUs. That would mean 508 points for Zen 4 - that is still slower than Alder lake.

There isn't a person alive that would buy any of those processors and run them single core 3.6 GHz locked. What's even funnier about your post is that the data doesn't show what the heat output or power consumption is during this test and I doubt seriously that the 12900K is the coolest or efficient at that speed.

 

[Kocicak]

The point of my post was to compare the processors running at the same speed, Alder and Raptor lake cores are/will be very strong cores not to be frowned upon.

AMDs saving grace now is the new very efficient TSMC 5 nm process. Their next big jump in IPC comes with the next CPUs, ZEN 4 is just an improved ZEN 3.

 

[therealmongo]

You really are fighting a losing battle. Why don't you just come out and say "mehhh AMD te suxx, Intel gud" already?

If we ignore all that high frequency copper melting inferno happening now and compare the single core output at nice cool enviromentally friendly 3600 MHz, we get this:

View attachment 66900

If I remember correctly, AMD stated that the new CPUs are 9% better at Cinebench than 5000 series CPUs. That would mean 508 points for Zen 4 - that is still slower than Alder lake.

Please, please, dont buy one of the Zen4 CPUs, we dont want to have to go through you moaning you didnt get a good one and sending them all back continuously until you get a better one

 

[Kocicak]

I need to get something else because my Alder lake computer keeps freezing. I do not believe I will be able to solve it, I am not an expert.

I was pretty happy with 5600X and discrete GPU before I got the idea of simplifying things by running 5700G, which my motherboard did not like at all. While solving the mess I thought I might as well try Alder lake.

 

[Makaveli]

If we ignore all that high frequency copper melting inferno happening now and compare the single core output at nice cool enviromentally friendly 3600 MHz, we get this:

View attachment 66900

If I remember correctly, AMD stated that the new CPUs are 9% better at Cinebench than 5000 series CPUs. That would mean 508 points for Zen 4 - that is still slower than Alder lake.

man that sucks for all the Pro Cinebench gamers out there what are they going to do in their leagues now

 

[In2Photos]

The point of my post was to compare the processors running at the same speed, Alder and Raptor lake cores are/will be very strong cores not to be frowned upon.

Was it? It sounded like you just wanted to throw shade at temps and efficiency like you have done in the past.

If we ignore all that high frequency copper melting inferno happening now and compare the single core output at nice cool enviromentally friendly...

AMDs saving grace now is the new very efficient TSMC 5 nm process. Their next big jump in IPC comes with the next CPUs, ZEN 4 is just an improved ZEN 3.

Their "saving grace"? Why not call it innovation? What has Intel done in the last few years regarding IPC? We don't have to look too long ago at the disaster of 11th Gen and the joke that was. But ok, let's knock AMD on Zen 4 vs Zen 3 when there is clear evidence of improvement.

 

[Thunder 57]

Was it? It sounded like you just wanted to throw shade at temps and efficiency like you have done in the past.



Their "saving grace"? Why not call it innovation? What has Intel done in the last few years regarding IPC? We don't have to look too long ago at the disaster of 11th Gen and the joke that was. But ok, let's knock AMD on Zen 4 vs Zen 3 when there is clear evidence of improvement.

Remember, this guy wanted a 3600 x 2 for cheaper than a 3800X.

 

[Thunder 57]

Please, please, dont buy one of the Zen4 CPUs, we dont want to have to go through you moaning you didnt get a good one and sending them all back continuously until you get a better one

Oh yeah, he also did that. What a bastard.

 

[Det0x]

Here some data from me for GB 5.4.5 on Linux with 65W(88W) ECO mode (I always have this on always)

Gigabyte Technology Co., Ltd. X570 AORUS PRO - Geekbench

Benchmark results for a Gigabyte Technology Co., Ltd. X570 AORUS PRO with an AMD Ryzen 9 5950X processor.

browser.geekbench.com

Above is all core load.
If we get 50% perf at same power that is insane or <50% lesser power with same score.
Dev Laptops with Dragon Range are just going to be insane. Better than a stock 5950X.

Our numbers match pretty good despite me running windows. 😊

My 5950x running tweaked win10 @ same eco 65w (88ppt) power settings.


Mem settings used which i guess boosted my scores alittle even with me running windows (should be slower than linux)

Edit
Forgot to write that i had pbo co disabled and stock 5050mhz fmax

 

[nicalandia]

I've been scouting GB5 to see what I found and I am reminded these words.

"This is the most expensive AMD Ryzen CPU you can buy, which as you might suspect is also the fastest and most powerful Ryzen CPU currently in existence. Not even upcoming Zen 4 CPUs will change that"

Here is the King.

GIGABYTE W771-Z00-00 - Geekbench

Benchmark results for a GIGABYTE W771-Z00-00 with an AMD Ryzen Threadripper PRO 5995WX processor.

browser.geekbench.com

I mean King of Release and September Release CPUs. Because this guy Zen4 Genoa is just bullying anything that comes near GB...

 

[Carfax83]

What leaks? Haven't yet seen any hint at the max achievable fabric clock.

And in the presentation AMD had their Zen 4 rig running with DDR5 6000 so to me this implies the leak is accurate.

It will be just like Zen 3 in the sense that 1:1 ratio will be the best for performance. Running the memory at higher speeds than the FCLK will increase bandwidth but probably result in a latency penalty.

https://twitter.com/x/status/1557258531115606016

 

[Hans Gruber]

So 6000mhz for DDR5 right out the box maxes out the 1:1 fabric clock on Zen 4? DDR5 is only barely a year old. Does that mean DDR6 is right around the corner?

 

[moinmoin]

I'm kinda disappointed nobody seems to have really thought about what AMD talked about with regards to Zen 4C honestly.

Same ISA support, same IPC, and now also confirmed to use half of the core area (key word being core, mind you).

That puts Zen 4C in similar size regions to ARM cores such as V1. That's kind of a big deal.

I guess I missed it. Did Papermaster talk about that? Or is it somewhere else? Press mentions of Zen 4c seem to be limited to it using N4.

View attachment 66897

That looks nice! What's the CLI tool you used to print these? Thanks.