Why everyone say Zen 5 is bad ?

[igormp]

That's putting a lot of hope in a rumor. With Intel cutting cost in a lot of places, I'd be surprised to see them keeping this. This isn't the obvious time for them to have (almost) two new CPU types at the same time.. or do they want LGA1700 to become their AM4?

What would we do without hope.

Isn't it also supposed to be an edge/networking solution at first? I believe that, even if it were to came to the desktop, it'd take a while for it to be available through the regular sales channels.

 

[SL2]

Isn't it also supposed to be an edge/networking solution at first? I believe that, even if it were to came to the desktop, it'd take a while for it to be available through the regular sales channels.

I've heard that as well, which begs the question if it would run anywhere close to 14900K speeds.

Without the speed there's no point drooling at the core count. It's the same situation as when looking at server parts.

 

[Zach_01]

Heard some rumors that Bartlett Lake could suffer from the same issues like Raptor so maybe its meant only for low-er speeds...
But nothing is confirmed of course. Could be just bad mouthing, but I would keep my hopes low.

 

[SL2]

Heard some rumors that Bartlett Lake could suffer from the same issues like Raptor so maybe its meant only for low-er speeds...
But nothing is confirmed of course. Could be just bad mouthing, but I would keep my hopes low.

Yeah you never know, but unless a CPU is built with high clock speeds in mind, what are the chances it would end up as fast as people wants it to be?

Is it close to being a Raptor with swapped cores, or are there other changes like number of PCIE lanes? (doubt it)

What about RAM speed? Is the multiplier unlocked?

Now that I think of it, the chance of it having an unlocked multiplier alone seems unlikely.

 

[AusWolf]

You must self censor as a beginning, then we will talk, if you are still here..

I call bullshit on that, too. Let's not get personal, shall we?

 

[Wolverine2349]

It really did.

and they killed HEDT which sent all of us core whores down to consumer.

Yeah and AMD never had an HEDT with more than 8 cores on a single CCX within a single CCD. Even Threadripper Zen 3 and newer still 8 cores per CCX/CCD.

AMD HEDT is great if you want more than 16 cores and lots of them in separate NUMA domains.

Intel on other hand had HEDT with more than 8 cores a while back last being Cascade Lake X which had decent IPC. It was on a mesh, but all cores were on same mesh and overclocking the mesh yielded much better performance despite latency crippled compared to ring bus and sometimes the overclocking of mesh negated that latency penalty if I remember right?

Now Intel has no HEDT. They have Xeon Workstation Sapphire Rapids which is more expensive and its IPC is crippled unlike Skylake X and Cascade Lake X requires ECC RAM which is less overclock friendly and stupid glamorous enterprise grade motherboards detrimental to things like gaming. Though but they do have more than 8 strong cores on one die the mesh, but their IPC and latency is once again severely gimped compared to Alder Lake client Golden Cove.

 

[R-T-B]

That's wrong. Six cores only would not be enough for the game and Windows and all of its main and background processes to run without micro-stuttering, and drops of the 0.1% lows.

I love the evidence ARF brings to the table...

EDIT: in case it isn't clear, the post I quoted is lacking it.

 

[Wolverine2349]

Isn't it also supposed to be an edge/networking solution at first? I believe that, even if it were to came to the desktop, it'd take a while for it to be available through the regular sales channels.

Intel Bartlett Lake-S Desktop CPUs Launching In 2025: Up To 8+16 Hybrid & Up To 12 P-Core Only Flavors

Intel's Bartlett Lake-S Desktop CPUs are reportedly launching in 2025 in hybrid and P-Core-only flavors with up to 12 P-Cores.

wccftech.com

Looks like a consumer release? Though hard to say,

Though if its networking and edge, how much does it cost Intel to make a 12 + 0 die on a ring. Or would they just shave 12 working P cores off Sapphire Rapids or Emerald Rapids defective dies and stuff them in it?? If its the later a hard pass, If its the former I am probably a buyer.

Why would Intel go through hassle of making a new 12P core die on a ring bus for only networking and edge when ring bus is really for performance desktops?

Yeah you never know, but unless a CPU is built with high clock speeds in mind, what are the chances it would end up as fast as people wants it to be?

Is it close to being a Raptor with swapped cores, or are there other changes like number of PCIE lanes? (doubt it)

What about RAM speed? Is the multiplier unlocked?

Now that I think of it, the chance of it having an unlocked multiplier alone seems unlikely.

If it had just 4 more PCIe Gen 4 lanes awesome. Then could have 2 NVME off CPU without cutting into 16 Gen 5 video card lanes. I like 2 NVME SSDs. One for OS and one for games.

Lisa basically admitted in an interview a little while ago that AMD is sandbagging on core counts

Q&A with AMD CEO Dr. Lisa Su

Computex 2024 Press Roundtable

morethanmoore.substack.com

Well how about more than 8 cores on a single CCX that isn't C cores Lisa Su.

Per rumored even Zen 6 is still 8 regular cores and 16C cores and 32 denser cores on CCXs. That sucks.

So is she really saying they could put more than 8 on a CCX that are big cores and not the C cores??

Or just that maybe they would consider adding more to consumer chips as in 8 core CCX/CCDs, but maybe increase it to 3 instead of 2. No thanks. Threadripper already does that a whole bunch of 8 core CCDs.

AMD seems technically limited in their design unlike Intel and maybe struggles to get more than 8 on one die. Intel has more options and much bigger company. They just have not executed them lately or provided good ones to consumers or gimped IPC and latency and ECC required expensive SPR Xeon Workstation.

 

[oxrufiioxo]

AMD seems technically limited in their design unlike Intel and maybe struggles to get more than 8 on one die. Intel has more options and much bigger company. They just have not executed them lately or provided good ones to consumers or gimped IPC and latency and ECC required expensive SPR Xeon Workstation.

It's a cost thing and it's competitive enough with intel that they don't care to offer more. Seems the primary goal is to use 1 8 core CCD for the whole product stack which honestly has worked out pretty well for them.

They do make odd intel like chips for laptop though



They actually do seem to care more about server and laptop so it makes sense I guess.

 

[SL2]

So the ADMIN-Admin (or whatever you call it) is one thing, but using 24H2 for tests is not something I've seen mentioned before.

Moreover, AMD uses Admin mode for its automated testing on Windows 11 24H2. This mode enables new branch prediction code optimizations that were not present in the systems used by reviewers. These optimizations are only available in the latest version of Windows 11, which is currently accessible only through the Insider Program (Build 26100). In other words, gaming performance is expected to improve once the newer version is released.

AMD promises Windows 11 patch boosting Zen5, Zen 4, and Zen 3 gaming performance - VideoCardz.com

AMD clarifies discrepancies in gaming performance between official data and reviews The company says that Windows patch will improve gaming performance, explains why ‘admin’ account was used for internal testing. AMD published a blog post discussing the disparity between the official gaming...

videocardz.com

Ryzen 9000 Series Community Update: Gaming Performance

The new “Zen 5” architecture first introduced in high end laptops earlier this summer is now available in a full stack of desktop processors with the AMD Ryzen 9000 Series. And while “Zen 5” is delivering leadership performance across content creation, productivity and AI applications, we’ve...

community.amd.com

I just wish they would have told us from the beginning!

 

[oxrufiioxo]

So the ADMIN-Admin (or whatever you call it) is one thing, but using 24H2 for tests is not something I've seen mentioned before.

AMD promises Windows 11 patch boosting Zen5, Zen 4, and Zen 3 gaming performance - VideoCardz.com

AMD clarifies discrepancies in gaming performance between official data and reviews The company says that Windows patch will improve gaming performance, explains why ‘admin’ account was used for internal testing. AMD published a blog post discussing the disparity between the official gaming...

videocardz.com

Ryzen 9000 Series Community Update: Gaming Performance

The new “Zen 5” architecture first introduced in high end laptops earlier this summer is now available in a full stack of desktop processors with the AMD Ryzen 9000 Series. And while “Zen 5” is delivering leadership performance across content creation, productivity and AI applications, we’ve...

community.amd.com

I just wish they'd told us from the beginning!

At least now they are being honest and saying a 5-8% improvement in gaming in their cherry picked games using in game benchmarks ofc lmao......

Also that it's faster than Raptor lake if we artificially gimp the shite out of it lol....

 

[windwhirl]

using 24H2 for tests

I wonder who thought it was a great idea to use a Windows Insider build for testing.

That or the shiny new Enterprise LTSC

 

[SL2]

Also:

We're collaborating with Microsoft to roll out this optional update to all Windows 11 users soon.

Also that it's faster than Raptor lake if we artificially gimp the shite out of it lol....

Dunno if using "default settings" is AMD's fault. It's kind of in the name if you're asking me.

I wonder who thought it was a great idea to use a Windows Insider build for testing.

It's explained in the text. The problem is that AMD didn't have the update ready, or at least tell us.

 

[oxrufiioxo]

Dunno if using "default settings" is AMD's fault. It's kind of in the name if you're asking me.

Must have been how Tomshardware tested as well they are the only place that has the 7800X3D 30% faster than Raptorlake lol

 

[RUSerious]

I just wish they would have told us from the beginning!

100%. I don't know why so many hardware companies are being so dodgy lately. Who do they think they are? Apple?

 

[SL2]

100%. I don't know why so many hardware companies are being so dodgy lately. Who do they think they are? Apple?

AMD is only two weeks late tho. Damage is done, but at least it's not YEARS..

 

[oxrufiioxo]

AMD is only two weeks late tho. Damage is done, but at least it's not YEARS..

The problem is even when they were contacted by reviewers sayings what's up with these poor results they were dodgy about it.

 

[SL2]

Speaking of Toms, theres a driver coming soon as well, which is no surprise by now.

It's not mentioned in my links above tho (I think)

AMD’s blog doesn’t address a pressing issue with its chipset drivers, but I got clarification that the company will soon release a fix.

AMD's chipset driver uses a core parking technique to boost performance with some dual-CCD chips, like its new Ryzen 9 9950X and the top-tier Ryzen 7000X3D parts. However, this feature can’t be uninstalled from the operating system — once you’ve installed an AMD chip that uses the mechanism, you’re stuck with core parking for the life of your operating system. If you later install another processor with the same operating system, the feature will persist and can continue to park cores (potentially unbeknownst to the user), thus reducing performance with processors not designed to use the feature.

The only known good fix is to completely reinstall the operating system. This also creates problems for reviewers who test multiple processors on the same motherboard. While AMD didn’t touch on the topic in its blog, it’s conceivable that this issue contributed to the inconsistent results we’ve seen with the first wave of Ryzen 9000 reviews.

This issue has been known for some time, so I asked David McAfee, the Corporate VP and GM of AMD’s Client Channel Business, if there is a fix coming.

“[…] I will tell you it's in development. Our goal is to have that in your hands by the time we do our X3D reviews [Ryzen 9000X3D]. The timing didn't work out here, which was unfortunate, but I think this is a critical piece of ensuring consistent performance across a range of processor compares in the same socket,” McAfee responded.

AMD updates Zen 5 Ryzen 9000 benchmark comparisons to Intel chips — details 'Admin' boost coming to Windows 11, chipset driver fix

AMD tries to clear the air.

www.tomshardware.com

The problem is even when they were contacted by reviewers sayings what's up with these poor results they were dodgy about it.

They were? I've only heard what PCWorld-Gordon has said, which was pretty much the opposite. That's my only source tho.

To me it looks like they made a big mistake, and it took some time to figure it out.

 

[oxrufiioxo]

AMD updates Zen 5 Ryzen 9000 benchmark comparisons to Intel chips — details 'Admin' boost coming to Windows 11, chipset driver fix

AMD tries to clear the air.

www.tomshardware.com

They were? I've only heard what PCWorld-Gordon has said, which was pretty much the opposite. That's my only source tho.

To me it looks like they made a big mistake, and it took some time to figure it out.

Hub contacted them asking why performance was nowhere near what AMD claimed it was and intel basically told them your numbers aren't wrong but ours are higher. They then asked what settings were being used for the test systems etc and they wouldn't answer.

 

[SL2]

Hub contacted them asking why performance was nowhere near what AMD claimed it was and intel basically told them your numbers aren't wrong but ours are higher. They then asked what settings were being used for the test systems etc and they wouldn't answer.

Yeah, I remember now.

Must have been how Tomshardware tested as well they are the only place that has the 7800X3D 30% faster than Raptorlake lol

In games at 1080 I see a difference of 11% between 7800X3D and 14900K in the review of the latter, and 12% between the same in the 9950X review.

Sounds like too much, but I don't think "default settings" is to blame, maybe the reviewer..

 

[JWNoctis]

Hub contacted them asking why performance was nowhere near what AMD claimed it was and intel basically told them your numbers aren't wrong but ours are higher. They then asked what settings were being used for the test systems etc and they wouldn't answer.

For what it worth, there is a theory floating around that between the launch delay and the apparent non-utility by a single thread of the second decoder new in Zen 5, that there was some sort of last-minute microcode patch that fixed some bug or exploit, but regressed performance after the official benchmark was released. It certainly sounded interesting.

 

[oxrufiioxo]

Yeah, I remember now.


In games at 1080 I see a difference of 11% between 7800X3D and 14900K in the review of the latter, and 12% between the same in the 9950X review.

Sounds like too much, but I don't think "default settings" is to blame, maybe the reviewer..

In their original 9700X review they had the 7800X3D 30% ahead of the 14700k when typically it's around 10%... Even when hub used the 125w bios on the 14900k it was only 8% slower than the 7800X3D.

Obviously a reviewer can skew results however they want but in you just look at multiple reviews that don't artificially gimp intel that is about ballpark.

 

[SL2]

In their original 9700X review they had the 7800X3D 30% ahead of the 14700k when typically it's around 10%...

They must have changed it, it's 17% now, and 14% in the 14700K review. I don't believe any of those numbers anyway, they're too high compared to others.

 

[oxrufiioxo]

They must have changed it, it's 17% now, and 14% in the 14700K review. I don't believe any of those numbers anyway, they're too high compared to others.

I haven't followed toms in ages they use to be my go to back in like 2008/09....

About the time I was like nah this site sucks now.....

https://www.reddit.com/r/nvidia/comments/9a4jnw

 

[SL2]

I haven't fallowed toms in ages they use to be my go to back in like 2008/09....

About the time I was like nah this site sucks now.....

I never really liked them that much to begin with, I lost interest around 2004. AT, Xbitlabs and SPCR had lot more details in reviews, even if not having much in common.