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!

 

[inf64]

Redgamingtech claims in his latest YT video that he knows how is 7950X boosting depending on the load (take with grain of salt of course):



This implies that 7700X should be boosting to ~5.1-5.3Ghz on all threads.

 

[Asterox]

Redgamingtech claims in his latest YT video that he knows how is 7950X boosting depending on the load (take with grain of salt of course):
View attachment 67297


This implies that 7700X should be boosting to ~5.1-5.3Ghz on all threads.

That is not news, it is expected for 16 Core Zen 4+TSMC5nm+much higher CPU frequencies+higher TDP.

Cinebench R23 or similar test, red is expected.

As far i see, any Zen 4 CPU will hit 5.1ghz all cores/Threads boost even R5 7600X.

 

[Timmah!]

Redgamingtech claims in his latest YT video that he knows how is 7950X boosting depending on the load (take with grain of salt of course):
View attachment 67297


This implies that 7700X should be boosting to ~5.1-5.3Ghz on all threads.

So 5,2GHz all-core boost and 5,4 when half the cores are loaded? Very nice, me gusta. That should keep stuff running fairly smooth even with some things running in background. Way better than my skylake-x, that would spike to 4,5 or 4,6GHz for second or 2, say in AutoCad, but right away drops down to 4 or less.

 

[Kaluan]

So it seems:
5,4
5,55
5,75(?)
5.85
Are the fmax clocks for 7600X/7700X/7900X/7950X?

 

[Zepp]

And that was a mess in mobile from the very beginning. Was 2000 Zen+? No it was Zen. Was 3000 Zen 2? No it was Zen+. 4000 finally cleaned it up until 5000 made it a mix of Zen 2 and 3, great! That's why having the architecture generation as a specific digit would/will be such an big improvement.

Yes this. I commented in another discussion on this that
3250u is zen1
3300u is zen+
3300X is zen2
4300U is zen2
5500u is zen2
that's how ridiculous they got with it. i don't like this new naming scheme but if they stick with it that will at least be some consistency.

 

[nicalandia]

3250u is zen1
3300u is zen+
3300X is zen2
4300U is zen2
5500u is zen2
that's how ridiculous they got with it.

You need to look it from the Entire Package point of view, there are more things like uArch generation for a CPU, the 6000 Ryzen Zen3+ is much more than a + sign when compared to Zen3 based 5000 SOC.

 

[nicalandia]

Latest Geekbench Run for the 7700X

ASUS System Product Name - Geekbench

Benchmark results for an ASUS System Product Name with an AMD Ryzen 7 7700X processor.

browser.geekbench.com

 

[FangBLade]

I thought AMD is in rear-view mirror: https://wccftech.com/intel-anticipa...ugh-2023-aims-to-re-establish-growth-by-2025/

 

[inf64]

I thought AMD is in rear-view mirror: https://wccftech.com/intel-anticipa...ugh-2023-aims-to-re-establish-growth-by-2025/

I would really like to see in which niche workloads will Sapphire Rapids outperform Genoa and Genoa-X. They lost the ST performance crown, they never had MT performance crown, they will hardly win perf./watt. By the time they get Arrow Lake AMD will have Zen 5 and the circle continues...

 

[Timmah!]

I would really like to see in which niche workloads will Sapphire Rapids outperform Genoa and Genoa-X. They lost the ST performance crown, they never had MT performance crown, they will hardly win perf./watt. By the time they get Arrow Lake AMD will have Zen 5 and the circle continues...

Maybe they will do a bigger leap with Arrow Lake and level things up. Similarly like AMD did with Zen 1.

 

[nicalandia]

I would really like to see in which niche workloads will Sapphire Rapids outperform Genoa and Genoa-X.

From what I have seen on Xeon Phi HBM and SPR-SP HBM HPC/AI can seen impressive performance boost. AVX512/AMX should see boost in orders of magnitude with the help of HBM2e.

 

[igor_kavinski]

the 6000 Ryzen Zen3+ is much more than a + sign when compared to Zen3 based 5000 SOC.

Agree. For once I don't hate them naming Zen 3+ the 6000 series because it does differ fairly (the most being LPDDR5-6400 support). I just wish they would release a 6980HX laptop with 128GB LPDDR5-6400 and deliver it to my doorstep. I will not say anything bad about them for the next, ummm....5 years!

AMD community manager, feel free to PM me with the good news

 

[igor_kavinski]

Maybe they will do a bigger leap with Arrow Lake and level things up. Similarly like AMD did with Zen 1.

Seriously, unless they sacrifice a blue eyed virgin blonde the moment she hits puberty, they are not getting lucky any time soon. Anyone hoping for an Intel resurgence should be aware that AMD isn't sitting still. Resting on their laurels is not something they can do because they have a lot more to lose than Intel. They will continue to maintain their architectural lead, unless TSMC falters really badly and Intel finally manages to get back on track on their process roadmap.

Intel does make great products still (Alder Lake is honestly a surprise for me, especially the P-cores) but AMD has shown that they can fight fire with fire while using less gas

 

[Vope45]

Seriously, unless they sacrifice a blue eyed virgin blonde the moment she hits puberty, they are not getting lucky any time soon. Anyone hoping for an Intel resurgence should be aware that AMD isn't sitting still. Resting on their laurels is not something they can do because they have a lot more to lose than Intel. They will continue to maintain their architectural lead, unless TSMC falters really badly and Intel finally manages to get back on track on their process roadmap.

Intel does make great products still (Alder Lake is honestly a surprise for me, especially the P-cores) but AMD has shown that they can fight fire with fire while using less gas

I don't know, pat seems to be really confident about nodes advantage around 2025+

 

[igor_kavinski]

I don't know, pat seems to be really confident about nodes advantage around 2025+

He also keeps seeing AMD in the rearview mirror. Dementia? Alzheimers? Too much of Raja Koduri's Indian curry and Hyderabadi biryani?

I do like to watch him being all animated and enthusiastic though. Maybe he can do an Intel Reality show where he goes around the Intel HQ showing what they are working on and introducing their key architectural heads. Should be a blast and worth at least a million views on Youtube.

 

[eek2121]

With official Raptor Lake power limits being raised to 253w, he may want to start focusing on the products and stop talking crap about the competition. I know the quote is old. It is just ridiculous that every new Intel chip consumes more and more power and perf/watt goes lower and lower.

Anyways, has AMD released any details related to overclocking or PBO improvements?

 

[igor_kavinski]

The amount of development time within the Zen era would be the same as the Bulldozer era by 2025.

Your obsession with Bulldozer makes me wish you would change your profile pic to an actual bulldozer with AMD printed on it in huge letters

 

[ElFenix]

That's actually a great change if they consistently stick to it.

This is going to work really well until they get to hybrid CPUs where there are multiple generations of Zen core involved.

 

[uzzi38]

Redgamingtech claims in his latest YT video that he knows how is 7950X boosting depending on the load (take with grain of salt of course):
View attachment 67297


This implies that 7700X should be boosting to ~5.1-5.3Ghz on all threads.

That is fairly accurate, providing you can keep the chip cool.

Note: I'm not talking about having to increase PPT, I mean just keeping at boosting at 230W.

 

[uzzi38]

With official Raptor Lake power limits being raised to 253w, he may want to start focusing on the products and stop talking crap about the competition. I know the quote is old. It is just ridiculous that every new Intel chip consumes more and more power and perf/watt goes lower and lower.

Anyways, has AMD released any details related to overclocking or PBO improvements?

Ooh, there's something absolutely hilarious I'm dying to talk about and it involves Curve Optimiser. Outside of that thing, I have no clue.

 

[igor_kavinski]

Ooh, there's something absolutely hilarious I'm dying to talk about and it involves Curve Optimiser.

Please do share

 

[uzzi38]

Please do share

Relax, it's not particularly important or game changing. Just hilarious.

 

[JohnTheHero]

I would really like to see in which niche workloads will Sapphire Rapids outperform Genoa and Genoa-X.

This is only guessing as neither has been actually released, but I would probably bet on Intel in networking and telecommunication - QuickAssist really helps for some loads, especially if you are power constrained (or aim at efficiency) as accelerators are usually burning less power for same job.

At the same time, who knows what kind of surprises in this area will Genoa bring. Also ARM based SoCs are taking their piece of pie, so it's gonna be hard for Intel in coming years.

 

[igor_kavinski]

Just hilarious.

I'm always down for hilarity!

 

[Saylick]

Some dude in Japan.

View attachment 67296

LMAO, not so subtle shade at Intel? AMD is playing (Japanese) chess, while Intel is playing checkers?