10% just out of a Windows build, impressive, still the 9700X managed to be even slower on average than 7700X, proving Zen 5 is dead for gaming at the current price. Can't wait to see the Intel benchmarks!
 
https://www.hardwareluxx.de/index.p...rsage-zen-5-mit-windows-11-24h2-getestet.html

This update pushes the 7800X3D even further ahead of the competition and its new-generation Zen5 siblings. I think Arrow lake is going to have to deliver a lot of performance to keep up with AMD this time.

**I've seen users reporting gains in all generations from Zen2 to the present. I'm almost inclined to upgrade to W11, but my focus is on 1440P...
 
Out of curiosity, do we know that this update does not boost performance for Zen3 as well?

I'm on 5800X3d and as much as I am *not* a fan of Windows 11, I'd upgrade if free performance was on the table.
 
If we take AMD's marketing, the 9700X was going to be just a little slower or a little faster than the 7800X3D. They also were saying there was a 10-15% increase in performance from 7700X... Is it possible that AMD's internal testing was as confused as everyone else by this performance increase? I mean if you think about that increase for the 9700X(Admin mode) probably does get it close to the 7800X3D (not on 24H2). Could they have not bothered retesting the 7000 series and really believed they were going to see the gains they advertised? Then really shocked when reviewers were saying it's barely different?

It seems improbable, maybe even impossible, but it would make sense of the marketing vs the reality. I just couldn't imagine them being so lazy and AMD is pretty notorious now for misleading benchmarks.
 
I'm stoked for X3D benchmarks, and am curious to see how other Zen 3 CPUs are impacted by this.
 
I'm interested to find out what kind of gains that Zen 3 / 4 Laptop CPUs can expect (if any).

As an owner of a Zen 4 Ryzen 7 laptop, I'll have to run some benchmarks before and after 24H2, and report my findings.

I'm also curious to see the performance improvement these Ryzen processors gain in 24H2 Windows 11 vs Windows 10. I'm expecting Windows 11 will be pulling ahead of Windows 10 from a performance standpoint, which is great news going forward.
 
I'm stoked for X3D benchmarks, and am curious to see how other Zen 3 CPUs are impacted by this.
I'm curious on Zen 3 non-X3D too, I upgraded from a 5800X to a 7800X3D because I was seeing more and more games where I was losing significant performance on my 4080. Now I have to wonder, would I have been okay waiting a little longer had this update rolled out first. Looks like 5800X3D is getting a nice boost.

Still excited to see that the 7800X3D investment is better than anticipated.
 
Out of curiosity, do we know that this update does not boost performance for Zen3 as well?

I'm on 5800X3d and as much as I am *not* a fan of Windows 11, I'd upgrade if free performance was on the table.
It does upgrade Zen 3 and from what I have seen the 5800X3D will get a nice boost. But, wait a few more days and many more benchmarks should be coming out.
 
Nature of the beast. Interesting to see Microsoft reasoning
How quickly they update for Nvidia or Intel. Could just be timelines AMD delivered on time . MS did not

I'm sure these CPUs will get better over time, Still wondering if A team designing the Zen 6 and not this

If game developers optimise for AMD - then AMD would probably get a 5% boost and Nvidia a 5% drop = that's the advantage of market dominance , it is what it is. Why I always said AMD should try to capture middle and low end.

STEVE - a call out, you won't do it, and pretty hard as apples to oranges, too hard to draw conclusions
PS5 AMD card vs same card running same games on PC , same settings. Once ripped from a PS5

or a hacked PS4 with windows/linux installed

Given that with modelling and statistical analysis you probably could compare what an AMD card in a PS5 could do on a PC - again see latest Commodore 64 games vs 1980 games or say Switch. Developers really learn how to get the most

Sometime in the future AI could assess how inefficient games are , compared to AI code 99% efficient
 
Curious if it has any effect on productivity… I’ve got the 7980x and will be testing that in the days to come…
 
Mindblowing! That much performance kept under the rug for all this time. Simply unlocked with a Windows update, wow.

Let's one guess how much performance is destroyed by imature drivers, bad game design and unoptimized OS. Consumers are constantly upgrading CPUs and GPUs because everybody else in this value chain of the gaming industry is focusing on fast bucks and fast rollouts. And every company basically cooks alone and the consumer (at least in PC Land) is forced to eat the conglomerate at home. Imagine the money saved if the software would be working just right with the hardware. We would be happy campers with old PC rigs, I.e. with Zen2 and RDNA2- Like those PS5 folks ;)
 
https://www.hardwareluxx.de/index.p...rsage-zen-5-mit-windows-11-24h2-getestet.html

This update pushes the 7800X3D even further ahead of the competition and its new-generation Zen5 siblings. I think Arrow lake is going to have to deliver a lot of performance to keep up with AMD this time.

**I've seen users reporting gains in all generations from Zen2 to the present. I'm almost inclined to upgrade to W11, but my focus is on 1440P...
Granted a decent boost on the Ryzen 5 7600X. Worthy of an architectural upgrade, it's insane when you think that Intel delivered +2-5% improvement with each "new architecture".
 
Kudos to Steve. Great review.
 
Insane that MS gimped AMD performance across the board by 10%!!!
 
Not impressive, Micro$oft. This only reveal you have been "teaming up" with that Chipzilla all along these years.
 
Curious to see if the 7800X3D also benefits from this update.
 
I suspect this issue is a remnant of the Spectre vulnerability mitigations. Spectre was using branch prediction to inject malicious code and some software mitigations were put in place. On later versions of the cpus these vulnerabilities were supposedly addressed at hardware level. Maybe the software mitigations were still in place needlessly affecting the branch prediction performance. I find it hard to believe that someone just found a magic trick to bring more performance to the table, but removing a known obstacle is more likely.
 
This is more in line with the Phoronix review. Zen5 is faster. Zen5 does it with less power. It's not a leap but a solid step forward. Just as AMD failed to work with Microsoft to offer this since day one (and its old, because incredibly it pushes a lot the performance of previous architectures too), journalism as to take it to the chin having failed to find this at review time, just spitting benchs without making the right questions of why is underwhelming.
 
Yes, very interesting, but please add for 1-2 games an 5700x3d or 5600x to the test!
 
So not to be too critical, but the article mentions a simple trick and how you are going to show us how it's done, but then dance around the specific steps needed. It's possible one or two paragraphs are just missing, but just in case...

Seems like a direct link to the patch and a statement saying "install this" would suffice, but providing a link to Microsoft's page about the 22631 optional update that doesn't even mention the AMD optimizations could lead to more confusion. Unless the simple trick is to upgrade Windows to the latest Insider Build version, which some may not consider simple?

Also, you say you tested 23H2 both with the patch and without, but then I didn't see anything indicating the difference of 23H2 with/without the patch - the only comparisons were of 24H2 and 23H2 (patched - and that was a preview build). Is there a difference between patched/non-patched or is that a different article?

Granted I'm reading this on mobile, so I may have missed it - but if you are making claims about showing the simple fix - then it might be more helpful to make it more obvious so folks don't have to search outside of techspot for the solution.

(Meanwhile, here's the link to 22H2 and 23H2 optional patch (KB5041587)).
 
Last edited:
Some of those improvements too good to believe especially considering these 24H2 changes are meant for CPUs which have a expanded branch prediction capacity aka Zen 5 so how come Zen 4 is getting better improvements. Puzzling
 
Well, that's fresh install time.
 
Could it be possible when you are seeing the massive gaming performance boost is because of an updated game version? Like when you are doing the test are they both the same versions of the game?
 
Mindblowing! That much performance kept under the rug for all this time. Simply unlocked with a Windows update, wow.

Let's one guess how much performance is destroyed by imature drivers, bad game design and unoptimized OS. Consumers are constantly upgrading CPUs and GPUs because everybody else in this value chain of the gaming industry is focusing on fast bucks and fast rollouts. And every company basically cooks alone and the consumer (at least in PC Land) is forced to eat the conglomerate at home. Imagine the money saved if the software would be working just right with the hardware. We would be happy campers with old PC rigs, I.e. with Zen2 and RDNA2- Like those PS5 folks ;)

Zen 2 launched 5 years ago and RDNA2 lauinched 4 years ago. Zen 2 is still perfectly viable today despite newer iterations of the Zen architecture running far better. Even RDNA2 is perfectly fine these days. It is Zen1 and older that are getting pretty long in the tooth by now along with the older Intel architectures. GPU wise you are usually fine with anything 10 series or better on Nvidia's side (the 900 series is lacking in VRAM) and anything RDNA or better on AMD.

In other words, my kids are happy campers with their Zen+ paired with a 2080 ti, a 4790 paired with a 980 ti (that kind of needs upgrading soon) and a 6800HQ paired with a GTX 1070.
 

Similar threads