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

[DisEnchantment]

Speculate at will

 

[Mahboi]

ARL
Arrow Lake
Arrow Lake
Arrow Lake Refresh

Might as well get to the point: there is no reason to expect anything out of Arrow Lake or its immediate successors, or anything from Intel for the next 2 years.

 

[CouncilorIrissa]

Not for your use cases perhaps, but for others. More specifically those that need max MT perf on DT at a reasonable price. Sounds like Arrow Lake Refresh will be a better option for them, if there will not be any Zen6 DT variant to counter it.

As if anything other than chewing through JS and games matters on desktop for 95% of the people.

 

[DrMrLordX]

Just a suggestion: using GNR as an abbreviation for Granite Ridge may not be a great idea, since it is also the abbreviation being used for Intel's Granite Rapids (a product that will exist concurrently with Granite Ridge, assuming Granite Rapids isn't canceled).

 

[Goop_reformed]

So you are suggesting 16C Zen6 will have better MT perf than Arrow Lake Refresh 8P+32E?

If not, then persons primarily needing max MT perf at reasonable price on DT should get Arrow Lake Refresh instead of Zen6 DT?

I think he's saying the single thread performance deficit is so great that those extra 16 e cores are meaningless.

i believe those zen 5 scores and if that’s what AMD is scoring at such a low clock then a >3Ghz base clock will be fun

Who knows? Scaling could be a problem.

 

[uzzi38]

Just a suggestion: using GNR as an abbreviation for Granite Ridge may not be a great idea, since it is also the abbreviation being used for Intel's Granite Rapids (a product that will exist concurrently with Granite Ridge, assuming Granite Rapids isn't canceled).

Don't forget Grand Ridge!

 

[DrMrLordX]

Don't forget Grand Ridge!

Ugh. Forgot about that one.

 

[Timorous]

Not for your use cases perhaps, but for others. More specifically those that need max MT perf on DT at a reasonable price. Sounds like Arrow Lake Refresh will be a better option for them, if there will not be any Zen6 DT variant to counter it.

Needs and wants are not the same.

Needing max MT performance on desktop at a reasonable price is an oxymoron. Either there is a genuine need in which case people are willing to pay quite a lot to fulfil it or it is simply a want and if the performance does not reach what you want at the price you are willing to pay then buy something that does or wait until something is available.

 

[Fjodor2001]

People that need nT have money.
For Threadrippers.

Those looking for max MT perf in the range that Arrow Lake Refresh 8P+32E provides will get it at a lower price by buying Arrow Lake Refresh, compared to if buying lower end Zen6 Threadrippers with same MT perf. Why pay more if you don't have to?

 

[Mahboi]

I think he's saying the single thread performance deficit is so great that those extra 16 e cores are meaningless.

Even if only in MT, who the heck expects the e cores to suddenly jump in 100% perf?
Last I looked, Gracemont is reaching in the ballpark of 60% of a Zen 2 core's perf. It's just not a big core.
You can add 100 of them if you like and make the "first client hundred core CPU" if you like, it's still a small core.

Jumping into Skymont is going to be a lot better, but 100% better? To the point that an e-core would reach above a Zen 2 core's perf? Let alone a Zen 5? Not likely.

 

[leoneazzurro]

Those looking for max MT perf in the range that Arrow Lake Refresh 8P+32E provides will get it at a lower price by buying Arrow Lake Refresh, compared to if buying lower end Zen6 Threadrippers with same MT perf. Why pay more if you don't have to?

Because with more cores at a certain point you'll need also other resources, i.e. memory bandwidth, i/o and so on which will be limited in a desktop platform. I also understand that quite probably nT performance on desktop, regardless of the vendor, will be limited mostly by thermals and power. So adding cores over a certain point wiithin a limited power window will add diminishing returns.

 

[Rekluse]

Why would it?
Why do people still have this weird fantasy that multicore is faster?
Multicore has diminishing returns for almost all applications. You don't need more than 8 for 95% of DT apps, don't need more than 16 for 99% of DT apps. The rare case of needing more is in server.

Under no illusions that, "MOAR CORES = FASTER" . Quite the opposite actually, the cores are fast enough as is. I'm just a weird edge case running a Proxmox VM server with core overprovisioning. Would love to have my VM's get all the real cores they need and run both my Linux and Windows Desktop VM's with 8 cores each, leaving the rest for my docker/TrueNas/Pfsense VM's

 

[Joe NYC]

Those looking for max MT perf in the range that Arrow Lake Refresh 8P+32E provides will get it at a lower price by buying Arrow Lake Refresh, compared to if buying lower end Zen6 Threadrippers with same MT perf. Why pay more if you don't have to?

If you want > 16 cores, the next increment would likely be 32 cores.

But then, if you start comparing 32 Zen 5 cores vs. 32 E-Cores, which is not a valid comparison, as far as comparing prices and performance.

16 core Zen 5 is likely going to be closer in MT performance to 8+32 ARL than 32 core Threadripper, which is also going to have more memory channels...

 

[Timorous]

Those looking for max MT perf in the range that Arrow Lake Refresh 8P+32E provides will get it at a lower price by buying Arrow Lake Refresh, compared to if buying lower end Zen6 Threadrippers with same MT perf. Why pay more if you don't have to?

Will it though? Maybe in CB MT it might do better than 16c 32t Zen but in workloads that don't use all the cores I suspect Zen 5 will be quite a bit better. Take something like Puget bench and I think Zen 5 will be faster in a lot of those tests and be generally better for people who have a mix of all core / some core and single core applications in their workflow.

 

[uzzi38]

Second GB5 score now as well

AMD BIRMANPLUS - Geekbench

Benchmark results for an AMD BIRMANPLUS with an AMD Eng Sample: 100-000000994-14_N processor.

browser.geekbench.com

Not sure if it's even real though tbh.

 

[CouncilorIrissa]

Those looking for max MT perf in the range that Arrow Lake Refresh 8P+32E provides will get it at a lower price by buying Arrow Lake Refresh, compared to if buying lower end Zen6 Threadrippers with same MT perf. Why pay more if you don't have to?

A niche within a niche (desktop DIY market). A total of 5 people care.

People mostly buy top-end parts because they also happen to clock the highest (i.e. have the highest ST / longest FPS bar charts wooooo), not because they need all of the cores. If ARL is significantly behind on ST, having additional E-cores won't change much.

 

[FlameTail]

Zen 5?

 

[Glo.]

That score is LMAO.

 

[Mahboi]

I thought Lenovo was the Leak-a-more company.
ASUS is getting ahead of themselves, before they used to overvolt and cook 7800x3Ds, now they straight up burn their relationship with AMD before they're even released.

 

[coercitiv]

Zen 5?

Yeah, coming to you from the creators of Inspect Element.

 

[deasd]

I see your demand of hearing more Zen5 support from mobo manufacturer and here is Gigabyte's:

We’ve got a hot scoop from Gigabyte this morning: the next-gen desktop Ryzen processor line is officially named Ryzen 9000. The ongoing speculation about whether AMD would adhere to its mobile naming convention and opt for “8000” for SKUs debuting in 2024 appears to be settled.

Gigabyte has now confirmed that their motherboards are also ready for the new series. Gigabyte has spoiled AMD’s surprise by confirming its 9000 series after all

AMD "Ryzen 9000" Zen5 desktop CPUs now confirmed by Gigabyte - VideoCardz.com

It’s Ryzen 9000, Gigabyte confirms The cat is out of the bag. We’ve got a hot scoop from Gigabyte this morning: the next-gen desktop Ryzen processor line is officially named Ryzen 9000. The ongoing speculation about whether AMD would adhere to its mobile naming convention and opt for “8000” for...

videocardz.com

 

[Timorous]

I see your demand of hearing more Zen5 support from mobo manufacturer and here is Gigabyte's take:

AMD "Ryzen 9000" Zen5 desktop CPUs now confirmed by Gigabyte - VideoCardz.com

It’s Ryzen 9000, Gigabyte confirms The cat is out of the bag. We’ve got a hot scoop from Gigabyte this morning: the next-gen desktop Ryzen processor line is officially named Ryzen 9000. The ongoing speculation about whether AMD would adhere to its mobile naming convention and opt for “8000” for...

videocardz.com

The naming convention has for desktop has always been odd numbers for zen iterations.

Zen is 1000
Zen 2 is 3000
Zen 3 is 5000
Zen 4 is 7000
Zen 5 looks to be 9000.

The only exception was that there was a Zen + which was 2000 and then APUs / OEM only / Embedded / Pro parts that were 4000/6000/8000.

 

[FlameTail]

So is Strix Point/Strix Halo/Kraken Point 8000 series or 9000 series?

 

[Tuna-Fish]

By AMD's current naming scheme, any mobile products released this year should be the 8000 series.

So of course they will be the 9000 series, lol. (I have no clue what they will actually be.)

 

[Det0x]

Zen 5?
View attachment 97709

Comparison against my ranked#1 7950X on hwbot
(my 7950X was running ~5.65ghz static allcore if i remember correctly)

Seem out of this world...

*edit*
Its a fake, "11ghz clockspeed"

 

[Ghostsonplanets]

So is Strix Point/Strix Halo/Kraken Point 8000 series or 9000 series?

All 9000. Same for Granite Ridge on Desktop.