Is it true that 7950X3D suffers from stutter on games?

Please, let me know how well your system runs without graphics card drivers. They should just work with the OS and DirectX or OpenGL, right? They shouldn't ever need to be updated to resolve performance issues.

That's how ridiculous your argument is. You're making a mountain out of a molehill with regards to latency and performance.
If after all this back and forth, you still don't understand that we don't want an EXTRA piece of software dictating core prioritization (that has not been needed in the past) then you're simply too dense to grasp anything beyond your limited bubble. Like I said, i'm sure if/when a 16 core vcach CPU becomes available, you'll strive for one with half the cores, based on your brilliant arguments here.
 
I said there's nothing else to troubleshoot. Are chipset drivers installed and game bar enabled. Lots of work there I guess, way more than building 2 systems.
There is something ELSE to troubleshoot. I've never needed xbox game bar and if I get a Frankenstein CPU, i'll need xbox gamebar.
 
You mean performance issues that existed and has since been resolved? Performance issues that has come up time and time again ever since the advent of Hyperthreading and logical cores? When Hyperthreading and Bulldozer were released, there was a need for drivers to tell the scheduler to fill independent cores/modules before scheduling things on the same core/module. That was before Microsoft baked it into the OS. You're speaking as if this is a problem unique to the 7950X3D but it is not.
You’re oversimplifying things.

Install Windows 11 - most CPUs just work. Use 7950X3D - addl software required. That software must be functioning correctly for optimal performance - requiring monitoring.

There’s no FUD - these are facts. 7800X3D doesn’t need it. 7950X doesn’t need it.
 
A pretty good use case for the 7950X3D would be installing a hypervisor and giving the 3d cache ccd to windows and then giving the higher boosting non cache ccd to linux.
Then setup passthrough or looking glass or just use a kvm to swap between windows and linux. I bought mine with that in mind, just been too busy to set it up.
 
You’re oversimplifying things.

Install Windows 11 - most CPUs just work. Use 7950X3D - addl software required. That software must be functioning correctly for optimal performance - requiring monitoring.

There’s no FUD - these are facts. 7800X3D doesn’t need it. 7950X doesn’t need it.
The chipset driver? because the game bar is essentially part of windows.
 
The chipset driver? because the game bar is essentially part of windows.
- Game bar
- Clean install of Windows
- AMD drivers

I've personally had my 7950X3D not working because of those components not being in harmony.

I don't understand why people are white knighting the sub-optimal setup the 7950X3D has. It's good...but all that I am saying is I'd rather have the 7800X3D for vcache (gaming) or the 7950X for productivity.

If it works for you and you are willing to deal with monitoring that - fantastic. I personally do not recommend it for the price.
 
If after all this back and forth, you still don't understand that we don't want an EXTRA piece of software dictating core prioritization (that has not been needed in the past) then you're simply too dense to grasp anything beyond your limited bubble. Like I said, i'm sure if/when a 16 core vcach CPU becomes available, you'll strive for one with half the cores, based on your brilliant arguments here.
Xbox game bar- already included by default in Windows 10 and 11 and is not uninstallable without extensive digging.

AMD chipset drivers- already installed to make your system function properly.

Tell me where the extra is.

I would not recommend anyone to buy a 16-core v-cache CPU because there is absolutely no gaming benefit and a productivity hit in performance. Simple facts.
You’re oversimplifying things.

Install Windows 11 - most CPUs just work. Use 7950X3D - addl software required. That software must be functioning correctly for optimal performance - requiring monitoring.

There’s no FUD - these are facts. 7800X3D doesn’t need it. 7950X doesn’t need it.
Actually both the 7950X and 7800X3D require the drivers to run optimally. Boost, sleep states, etc all require the drivers telling the OS how to interact and use the CPU.

In any case, my point is that there has been precedent in the past for use of drivers to optimize core usage and people for the most part had no issues with it.
- Game bar
- Clean install of Windows
- AMD drivers

I've personally had my 7950X3D not working because of those components not being in harmony.

I don't understand why people are white knighting the sub-optimal setup the 7950X3D has. It's good...but all that I am saying is I'd rather have the 7800X3D for vcache (gaming) or the 7950X for productivity.

If it works for you and you are willing to deal with monitoring that - fantastic. I personally do not recommend it for the price.
This is the exact same idiotic argument that my uncles use against Mercedes. They had one single experience 20 years ago with a high end model that was expensive to fix and write off the entire Mercedes brand as unreliable and expensive pieces of junk. Just because there were initial teething issues does not mean those issues are still relevant today.
 
There is virtually no downside to the 7950X3D, compared to the 7950x.

Techpowerup has the 7950x at 0.3% better on average, in their applications suite. That's......nothing.
If you focus on major multicore scores, such as Blender, Cinibench, Corona: The 7950x fairs a small bit better. But, I think you'd be hard pressed to feel the difference of 1 - 3 seconds. Especially considering that...

.....7950X3D uses 88 watts less in Blender multicore

38 watts less, average over their 47 application suite.

7950X3D is 14% better on average, in gaming at 1080p and 1440p, and 4k. That's more than a generational improvement. (and may be even better, after the chipset driver and bios improvements, since launch). And as we know, certain games are WAY better with the extra cache.

26 watts less, averaged over 13 games. And the 7950x is already pretty efficient, there.


The chipset driver, bios, and xbox gamebar situation is much improved since launch. And if you really need to, Process Lasso is a pretty easy extra step to virtually eliminate any possible issue with the scheduling for gaming. Again, you can simply set your steam folder so that anything which launches from there, automatically is "lasso'd" to the Vcache CCX.
 
Last edited:
Xbox game bar- already included by default in Windows 10 and 11 and is not uninstallable without extensive digging.

AMD chipset drivers- already installed to make your system function properly.

Tell me where the extra is.

I would not recommend anyone to buy a 16-core v-cache CPU because there is absolutely no gaming benefit and a productivity hit in performance. Simple facts.

Actually both the 7950X and 7800X3D require the drivers to run optimally. Boost, sleep states, etc all require the drivers telling the OS how to interact and use the CPU.

In any case, my point is that there has been precedent in the past for use of drivers to optimize core usage and people for the most part had no issues with it.

This is the exact same idiotic argument that my uncles use against Mercedes. They had one single experience 20 years ago with a high end model that was expensive to fix and write off the entire Mercedes brand as unreliable and expensive pieces of junk. Just because there were initial teething issues does not mean those issues are still relevant today.

I agree. 7800X3D for gaming. 7950X for productivity. 7950X3D if you want to have to monitor to ensure things are operating optimally.

Game bar - has to know the game in its database.
Clean Windows install - Windows is especially resilient in 2024 there is no need to do clean installs - unless you run a 7950X3D.
AMD drivers - installation issues will lead to problems - such as certain components not being recognized as necessary

How am I saying anything controversial? It's simply 7800X3D is the answer for gaming - better performance and value. 7950X is better for productivity, better value. I would not buy a 7950X3D again.
 
Hardware works thanks to the software, your point is pointless.
Yes, when the special, additional software that other CPUs do not require is working.

BTW - I have stutters in games with my 7950X3D - not with any of my 7800X3D rigs - back on topic of this thread.
 
- Game bar
- Clean install of Windows
- AMD drivers

I've personally had my 7950X3D not working because of those components not being in harmony.

I don't understand why people are white knighting the sub-optimal setup the 7950X3D has. It's good...but all that I am saying is I'd rather have the 7800X3D for vcache (gaming) or the 7950X for productivity.

If it works for you and you are willing to deal with monitoring that - fantastic. I personally do not recommend it for the price.
I never had a single problem on my 7950X3D, you are not able to use a PC, people like you needs a console not a PC.
I sincerely doubt that you have ever had a 7950X3D, you look like a troll trolling a product.

Never heard others complaining about the 7950X3D like you.
I'm a power user, I code on it in various languages, I use VMs, I render heavy 3d models on it (for my wife), I game on it, I never had a single problem if not an incredible feel of watching a rocket consuming so few with fans so silent.
 
Yes, when the special, additional software that other CPUs do not require is working.

BTW - I have stutters in games with my 7950X3D - not with any of my 7800X3D rigs - back on topic of this thread.
Shows us your stutter, please put your nickname in the afterburner osd.
I'm pretty sure that you are a kid that never seen a 7950X3D.
 
Xbox game bar- already included by default in Windows 10 and 11 and is not uninstallable without extensive digging.

AMD chipset drivers- already installed to make your system function properly.

Tell me where the extra is.

I would not recommend anyone to buy a 16-core v-cache CPU because there is absolutely no gaming benefit and a productivity hit in performance. Simple facts.

Actually both the 7950X and 7800X3D require the drivers to run optimally. Boost, sleep states, etc all require the drivers telling the OS how to interact and use the CPU.

In any case, my point is that there has been precedent in the past for use of drivers to optimize core usage and people for the most part had no issues with it.

This is the exact same idiotic argument that my uncles use against Mercedes. They had one single experience 20 years ago with a high end model that was expensive to fix and write off the entire Mercedes brand as unreliable and expensive pieces of junk. Just because there were initial teething issues does not mean those issues are still relevant today.
Unfortunately I never said I don't want to download an extra piece of software, I said I don't want to rely on it for proper thread assignment. So whatever point you think you made by telling us what we already know (that game bar is included) proved absolutely nothing except for maybe you're still not getting our beef.
 
Shows us your stutter, please put your nickname in the afterburner osd.
I'm pretty sure that you are a kid that never seen a 7950X3D.
I'd rather be a kid who's never seen a 7950x3d than an adult who's feelings are hurt that internet strangers aren't interested in a 7950x3d. Cry more buddy.
 
Software, software...

Is anything actually known about what the AMD "drivers" do?

We have little enough information about Intel's thread director. I have seen even less about AMD's hacks. I recall only that they communicate to the OS with a description of what core is what, and that only once when starting the OS.
 
I get that people can be passionate about things. Yet lets keep the name calling out of it (I know, I've been guilty in the past).

Recently I had some really bad frame pacing issues in a CPU intensive game, Star Citizen. It was the darndest thing - most of the time everything felt just fine, even in cities where performance is usually quite a bit lower. Yet in an asteroid cluster, I'd get several half second pauses and some WICKED frame pacing issues when flying into the cluster or just sitting still and rotating the ship around.

I did just about everything troubleshooting it... clean AMD chipset driver install, clean GPU driver install, trying different SSDs where I played SC from (including freshly downloaded game copies), SFC /scannow and DISM online checks, registry cleaning, and likely a bunch of other stuff that I forgot about.

It turned out all that I needed was a windows reinstall. Shitty windows screwing stuff up again. As near as I can tell, the likely culprit was a physical TPM failure that borked up windows in such a way that only a reinstall fixed whatever the underlying issue was.

TLDR - if you're having frame pacing issues (or as some would call it, hitching, or "stuttering" as if it were an audio problem) - try a windows reinstall. This was on my 7800X3D 4090 rig.
 
Last edited:
So let me get this straight - you want Intel and AMD to start releasing CPU drivers to address things like vcache on only half the cores? You’re so determined to defend the setup of the 7950X3D you won’t admit to the potential performance issues?

You’re not comparing apples to apples. CPUs should just work.

They do, what are you on about? It's part of the chipset drivers, you have to load them for a 7800x3d system to work optimally, too. What potential performance issues? It outperforms a 7950X in games no matter what.
There is something ELSE to troubleshoot. I've never needed xbox game bar and if I get a Frankenstein CPU, i'll need xbox gamebar.
Xbox game bar is installed and enabled by default, so as long as you don't fuck with windows, it works. This argument keeps getting dumber. The 7950x3d equals or beats a 7800x3d in gaming and matches a 7950x in productivity. If you need both, it's perfect. If you only care about one or the other, then get the one that matches your use case and save some money. The whole software/troubleshooting thing is BS in the extreme. And no, you don't need a clean windows install, mine is years old and was a win 10 upgrade and works perfectly. So you need to install AMD chipset drivers and that's it. You also need to do that to get the most out of a 7800x3d, so no extra work or software. But if you just game, get a 7800x3d. The 7950x3d is absolutely a waste of money for gaming. If you do productivity and game on the side, get a 7950x, It's absolutely a waste of money to get a 7950x3d. If you want the maximum from both then that's the only time it makes any sense to get a 7950x3d. And it's really still kind of dumb unless you competitively game. But I do, so it was worth spending some extra. I'm not stupid enough to buy both a7800x3d and a 7950x, though, that's just dumb. And I'm known to buy stupid stuff, just not that stupid.
 
Also boo hoo if the software fails to detect a game… you can always turn the other ccd off when it’s time to game. That’s what I did early on when I was playing a lot of Star citizen.

16 cores for renderings and simulations… then a bios toggle and you have a higher clocked 7800x3d for games.
 
Also boo hoo if the software fails to detect a game… you can always turn the other ccd off when it’s time to game. That’s what I did early on when I was playing a lot of Star citizen.

16 cores for renderings and simulations… then a bios toggle and you have a higher clocked 7800x3d for games.

So constantly reboot your computer any time you do something different? That sounds super convenient.
 
I don’t know why we are still going round and round.

Stutter happens with the 7950X3D.

https://www.google.com/search?q=7950x3d+stutter+reddit

Plenty of cases to review - all due to improper core utilization (CCD with vcache vs CCD with no vcache).

Sure, and 90% are 6+months old. Just like the start of this thread was.

And did you also search this?
https://www.google.com/search?q=7800x3d+stutter+reddit

Again, mostly FUD in here. 7800x3d if you game, 7950X if you work, 7950x3d if you do both and want minimal compromises. If you're going to whine about software tweaks, you probably shouldn't be on a PC. Get a Mac or a console. They 'just work' (except when they don't).
 
All this pissing and moaning in here for what? Does anyone here actually have experience with BOTH a 7800X3D and a 7950X3D? You are all going on and on about this, that and the other and linking to outdated reddit posts and other outside sources to validate each of whoever's point. That in and of itself is pointless because no matter what, if you google search a problem you will find people complaining about that specific problem. Shocking I know..... The point I'm trying to make is there are going to be people experiencing issues with virtually every piece of hardware somewhere because that's just what happens when you have myriad different combinations of hardware and software that it is impossible for any company to trouble shoot and validate their products on all combinations. So linking to threads as your "ah ha" moment or as definitive proof is useless. They can all be proven false just as quickly.

games tests are margin of error difference for me:
https://www.3dmark.com/compare/sw/903146/sw/942247

I however DO have experience with both a 7800X3D and a 7950X3D and have owned booth. As a first hand user of both products I have had no issue with either to be honest. Does that mean anything as a whole or as definitive proof one way or the other? Of course not but I can tell you in games they perform virtually identical to the point you'd need to be monitoring a frame graph to tell a difference. As for the argument of "needing" Xbox Bar being a hassle...well I got bad news for that individual......you've been using it in Windows this whole time as it's a core feature of the OS. It isn't "this whole other step" you have to do before playing anything. It's a completely invisible process, it's just on.

Someone mentioned about "needing to worry about drivers" and I'm not sure if that person is serious or just trolling. Ummm, yeah..... drivers are kind of a big deal.....or do you just raw dog your OS install and let Windows pull down whatever drivers it wants and then hope everything works? If you're that kind of person.....well there's this other operating system from a fruit company that may be more suited for you.
 
All this pissing and moaning in here for what? Does anyone here actually have experience with BOTH a 7800X3D and a 7950X3D? You are all going on and on about this, that and the other and linking to outdated reddit posts and other outside sources to validate each of whoever's point. That in and of itself is pointless because no matter what, if you google search a problem you will find people complaining about that specific problem. Shocking I know..... The point I'm trying to make is there are going to be people experiencing issues with virtually every piece of hardware somewhere because that's just what happens when you have myriad different combinations of hardware and software that it is impossible for any company to trouble shoot and validate their products on all combinations. So linking to threads as your "ah ha" moment or as definitive proof is useless. They can all be proven false just as quickly.

games tests are margin of error difference for me:
https://www.3dmark.com/compare/sw/903146/sw/942247

I however DO have experience with both a 7800X3D and a 7950X3D and have owned booth. As a first hand user of both products I have had no issue with either to be honest. Does that mean anything as a whole or as definitive proof one way or the other? Of course not but I can tell you in games they perform virtually identical to the point you'd need to be monitoring a frame graph to tell a difference. As for the argument of "needing" Xbox Bar being a hassle...well I got bad news for that individual......you've been using it in Windows this whole time as it's a core feature of the OS. It isn't "this whole other step" you have to do before playing anything. It's a completely invisible process, it's just on.

Someone mentioned about "needing to worry about drivers" and I'm not sure if that person is serious or just trolling. Ummm, yeah..... drivers are kind of a big deal.....or do you just raw dog your OS install and let Windows pull down whatever drivers it wants and then hope everything works? If you're that kind of person.....well there's this other operating system from a fruit company that may be more suited for you.
Just adding in that I have a 7950X3D rig and 3x 7800X3D rigs - the 7950X3D has stuttering to this day. Same game, same RAM, often the same GPU.

The 7950X3D requires more to work optimally than the 7800X3D. The end.

(Cue dozens of folks defending their CPU like it’s their child for some odd reason)

EDIT: receipts since the fanboys are rabid:

7950X3D - https://valid.x86.fr/25li5y
7800X3D #1 - https://valid.x86.fr/b8thxc
7800X3D #2 - https://valid.x86.fr/z70yul
7800X3D #3 - (waiting on GPU - used Nitro+ 7900 XTX coming Monday; Noctua 4080 S pre-ordered - will likely sell 4080 S FE)
IMG_5433.jpeg
 
Last edited:
Just adding in that I have a 7950X3D rig and 3x 7800X3D rigs - the 7950X3D has stuttering to this day. Same game, same RAM, often the same GPU.

The 7950X3D requires more to work optimally than the 7800X3D. The end.

(Cue dozens of folks defending their CPU like it’s their child for some odd reason)
Having them in 2 different systems is meaningless. Try them on the SAME system if you want a 1-1 comparison. I'm not defending anything here but you seem to be on some holy crusade against the 7950X3D and you're justifying with your own self indulging, self validating nonsense. I could sit here and invalidate all your claims as for ME there are no issues, there are no extra steps I run the game and it works, no stuttering, no irregularities no extra hassle. I COULD find and link to just as many threads proving my point that there is no issues. But I'm not and that's where you are missing the point of what I posted. YOU are not the "be all end all" of this topic, there are just as many people that aren't having issues with their CPUs for the same reason I posted above. Your "The end" comment is exactly what I'm talking about. You aren't "the end" your points aren't "definitive" nor are mine is what I'm saying here.

Instead of trying to attack and defend yourself or point, read the content and understand it first.
 
Having them in 2 different systems is meaningless. Try them on the SAME system if you want a 1-1 comparison. I'm not defending anything here but you seem to be on some holy crusade against the 7950X3D and you're justifying with your own self indulging, self validating nonsense. I could sit here and invalidate all your claims as for ME there are no issues, there are no extra steps I run the game and it works, no stuttering, no irregularities no extra hassle. I COULD find and link to just as many threads proving my point that there is no issues. But I'm not and that's where you are missing the point of what I posted. YOU are not the "be all end all" of this topic, there are just as many people that aren't having issues with their CPUs for the same reason I posted above. Your "The end" comment is exactly what I'm talking about. You aren't "the end" your points aren't "definitive" nor are mine is what I'm saying here.

Instead of trying to attack and defend yourself or point, read the content and understand it first.
I still don’t get your point.

The 7950X3D requires more software to work optimally than the 7800X3D and 7950X.

Do you argue against that? It’s a fact.

That fact means there’s more room for issues hence why there can be stuttering when switching between the CCDs (vcache versus no vcache).
 
OK have fun screaming into the void, but your facts are not facts at all. The same chipset and 3D v-cache optimizer drivers that are used for the 7800X3D are also used for the 7950X3D just as seamlessly and Xbox Game Bar is also used for both regardless but I'm not going to get sucked into your ranting nonsensical tirade. Have fun yelling at your screen but I'm going to go back to playing my problem free games.
 
Having them in 2 different systems is meaningless. Try them on the SAME system if you want a 1-1 comparison. I'm not defending anything here but you seem to be on some holy crusade against the 7950X3D and you're justifying with your own self indulging, self validating nonsense. I could sit here and invalidate all your claims as for ME there are no issues, there are no extra steps I run the game and it works, no stuttering, no irregularities no extra hassle. I COULD find and link to just as many threads proving my point that there is no issues. But I'm not and that's where you are missing the point of what I posted. YOU are not the "be all end all" of this topic, there are just as many people that aren't having issues with their CPUs for the same reason I posted above. Your "The end" comment is exactly what I'm talking about. You aren't "the end" your points aren't "definitive" nor are mine is what I'm saying here.

Instead of trying to attack and defend yourself or point, read the content and understand it first.
His experiences echo mine going from a 5900x to a 5800x3d, everything else is the same. A lot of micro stuttering disappeared, even in games that preferred clock speed. E.g. I get lower max fps in some games with the 5800x3D, but the game is no longer stuttering.

Problem is dual CCDs and mixed CCDs might make it worse. CPUs copy cache between cores to be able to switch the threads between them. When you have two CCDs you add another layer which creates large latency. The latency between CCDs is massive compared to going to Cache on the same CCD. E.g. core to core latency on the same CCD for the 5950x was sub 20ns on while it was in the 75-90ns range when between CCDs according to Anandtech on the 5950x. The 7950x is a few ns better, but still has the same issues.

AMD's dual CCD cpus are suboptimal for gaming, but not everyone notices the drawbacks and not all games have micro stuttering issues. Some get used to it and not notice it while others lack the perception ability to notice the micro stuttering. It is kind of like discussing monitors, some can't tell the difference between 60hz and 240hz, but for a lot of people that difference is massive.
 
His experiences echo mine going from a 5900x to a 5800x3d, everything else is the same. A lot of micro stuttering disappeared, even in games that preferred clock speed. E.g. I get lower max fps in some games with the 5800x3D, but the game is no longer stuttering.

Problem is dual CCDs and mixed CCDs might make it worse. CPUs copy cache between cores to be able to switch the threads between them. When you have two CCDs you add another layer which creates large latency. The latency between CCDs is massive compared to going to Cache on the same CCD. E.g. core to core latency on the same CCD for the 5950x was sub 20ns on while it was in the 75-90ns range when between CCDs according to Anandtech on the 5950x. The 7950x is a few ns better, but still has the same issues.

AMD's dual CCD cpus are suboptimal for gaming, but not everyone notices the drawbacks and not all games have micro stuttering issues. Some get used to it and not notice it while others lack the perception ability to notice the micro stuttering. It is kind of like discussing monitors, some can't tell the difference between 60hz and 240hz, but for a lot of people that difference is massive.
A silly question, yet I will ask it:

Isn't that exactly the reason why the Xbox Game Bar is used to keep game threads only on the X3D CCD? To prevent that cache spillover onto the non-X3D side?
 
A silly question, yet I will ask it:

Isn't that exactly the reason why the Xbox Game Bar is used to keep game threads only on the X3D CCD? To prevent that cache spillover onto the non-X3D side?
Based on the benchmarks I've seen of non-3D cache CCD disabled vs enabled indicate that it isn't working as ideally. If it is working perfect then the threads on the non-3D cache CCD should be pretty much idle. If it is working properly then the 0.1% lows on the 7950x3D should always be at least as good as the 0.1% lows on the 7800x3D.

When gamers nexus benched Counter-Strike 2 on the CPU the 7950x3D had 0.1 lows of 95fps while the 5600x had 172fps, the 7700x had 189 and 7800x3D had 203. Those 0.1% lows is a good indication of a lot of micro stuttering on the 7950x3D.

On the 5900x it would fill each core and then start hyperthreading so if you had 8 fully loaded cores then 2 would spill over to the second CCD as it preferred physical cores over logical cores. Wouldn't be surprised if it works kind of similar even with x3D CPUs in that it starts to use the second CCD when the first CCD is almost fully loaded on primary thread which then causes issues. In a perfect world you would just shut down the entire non-cache CCD on a hardware level when gaming is detected, but that would require a bit of work on both the hardware and the software side.
 
Based on the benchmarks I've seen of non-3D cache CCD disabled vs enabled indicate that it isn't working as ideally. If it is working perfect then the threads on the non-3D cache CCD should be pretty much idle. If it is working properly then the 0.1% lows on the 7950x3D should always be at least as good as the 0.1% lows on the 7800x3D.

When gamers nexus benched Counter-Strike 2 on the CPU the 7950x3D had 0.1 lows of 95fps while the 5600x had 172fps, the 7700x had 189 and 7800x3D had 203. Those 0.1% lows is a good indication of a lot of micro stuttering on the 7950x3D.

On the 5900x it would fill each core and then start hyperthreading so if you had 8 fully loaded cores then 2 would spill over to the second CCD as it preferred physical cores over logical cores. Wouldn't be surprised if it works kind of similar even with x3D CPUs in that it starts to use the second CCD when the first CCD is almost fully loaded on primary thread which then causes issues. In a perfect world you would just shut down the entire non-cache CCD on a hardware level when gaming is detected, but that would require a bit of work on both the hardware and the software side.
Again, a silly question:

With the updates over the past year (or so?) to the Xbox Game Bar and the AMD Chipset Drivers, has your source re-run those tests?

The most recent that I can find that digs down is from a year ago, and shows that when the Xbox Game Bar and Chipset Drivers are running properly, the non-X3D cores are fully parked:
View: https://youtu.be/9gCzXdLmjPY?t=392

Edit: same in the slightly newer 7800X3D review from them:
View: https://youtu.be/B31PwSpClk8?t=318
 
Last edited:
Again, a silly question:

With the updates over the past year (or so?) to the Xbox Game Bar and the AMD Chipset Drivers, has your source re-run those tests?

The most recent that I can find that digs down is from a year ago, and shows that when the Xbox Game Bar and Chipset Drivers are running properly, the non-X3D cores are fully parked:
View: https://youtu.be/9gCzXdLmjPY?t=392

Edit: same in the slightly newer 7800X3D review from them:
View: https://youtu.be/B31PwSpClk8?t=318

If you look at the 0.1% lows, which are always the best indication of micro stutters, they are fine in some of the games in the reviews you posted, but very bad compared to the rest of the field in others. Basically some games run fine while others have a good chance of micro stutter. People are so hung up on maximum average framerate as that is the e-penis++ number that they forget that the 0.1% lows and 1% lows matter much more for game experience. Havin 100fps average with 25fps 0.1% lows is not a good experience while 80fps average with 60fps 0.1% will be good for most as long as it is single player. In competitive multi player you generally want your 0.1% lows well above 100fps and average fps far north of 200fps in most engines.

Gamers nexus ran the CS2 tests in October of 2023 just after launch of CS2. CS2 had been 6 months in beta and it released around 7 months after 7950x3D release. It is a terrible sign if when it is not working properly after a game has been 6 months in beta and the CPU has been out for 7 months. Most games I play are 8-15 hours and I often buy at launch. It should work optimal at release.
 
If you look at the 0.1% lows, which are always the best indication of micro stutters, they are fine in some of the games in the reviews you posted, but very bad compared to the rest of the field in others. Basically some games run fine while others have a good chance of micro stutter. People are so hung up on maximum average framerate as that is the e-penis++ number that they forget that the 0.1% lows and 1% lows matter much more for game experience. Havin 100fps average with 25fps 0.1% lows is not a good experience while 80fps average with 60fps 0.1% will be good for most as long as it is single player. In competitive multi player you generally want your 0.1% lows well above 100fps and average fps far north of 200fps in most engines.

Gamers nexus ran the CS2 tests in October of 2023 just after launch of CS2. CS2 had been 6 months in beta and it released around 7 months after 7950x3D release. It is a terrible sign if when it is not working properly after a game has been 6 months in beta and the CPU has been out for 7 months. Most games I play are 8-15 hours and I often buy at launch. It should work optimal at release.
I can't find that info. Mind sharing it?

I did find a newer GN review (4 months old) that shows % lows on the 7950X3D and it's better than the 7800X3D or within margin or error:
View: https://youtu.be/2MvvCr-thM8

EDIT: This one? Steve talks about probably needing to use process lasso to properly assign the cores, so I'd assume that they didn't use game bar to make sure that it knows that CS2 is a game (and at the end of the update he talks about it being a game issue that they're working on):
View: https://youtu.be/aiqojNU-RBg?t=1197
 
Last edited:
I can't find that info. Mind sharing it?

I did find a newer GN review (4 months old) that shows % lows on the 7950X3D and it's better than the 7800X3D or within margin or error:
View: https://youtu.be/2MvvCr-thM8

EDIT: This one? Steve talks about probably needing to use process lasso to properly assign the cores, so I'd assume that they didn't use game bar to make sure that it knows that CS2 is a game (and at the end of the update he talks about it being a game issue that they're working on):
View: https://youtu.be/aiqojNU-RBg?t=1197

You could have gone to Gamers Nexus youtube page and searched for "Counter Strike" or "CS2" and found it right away.


View: https://www.youtube.com/watch?v=8mmeQ6DGIMY - tested with fps max 400.

View: https://www.youtube.com/watch?v=aiqojNU-RBg - retest of some CPUs that were bottlenecked by the max fps.
 
You could have gone to Gamers Nexus youtube page and searched for "Counter Strike" or "CS2" and found it right away.


View: https://www.youtube.com/watch?v=8mmeQ6DGIMY - tested with fps max 400.

View: https://www.youtube.com/watch?v=aiqojNU-RBg - retest of some CPUs that were bottlenecked by the max fps.

Yeah, that's the one that I linked...

"This one? Steve talks about probably needing to use process lasso to properly assign the cores, so I'd assume that they didn't use game bar to make sure that it knows that CS2 is a game (and at the end of the update he talks about it being a game issue that they're working on)" - them being the game developer, Valve.

Obviously it's a game issue in the case of CS2 for both E-cores and multi-CCD X3D CPUs, as acknowledged from Steve from what he has apparently heard from Valve.
 
Yeah, that's the one that I linked...

"This one? Steve talks about probably needing to use process lasso to properly assign the cores, so I'd assume that they didn't use game bar to make sure that it knows that CS2 is a game (and at the end of the update he talks about it being a game issue that they're working on)" - them being the game developer, Valve.

Obviously it's a game issue in the case of CS2 for both E-cores and multi-CCD X3D CPUs, as acknowledged from Steve from what he has apparently heard from Valve.
Works fine on all single CCD CPUs so it is just as much a CPU architecture issue as it is a game issue. Uniform cores on one CCD all work fine. The mixed core and mixed CCD CPUs are the outliers while others work fine. Personally I would only use one core complex CPUs with uniform cores for a gaming rig. E.g. disable e-cores on intel or only pick single CCD CPUs from AMD and rather build an extra machine if I need more cores for other stuff. The reason the 7800x3D was delayed was so that people wouldn't see it beating the 7950x3D on 7000 series x3D launch. Those two would be about equal or the 7950x3D would be slightly ahead if everything is working perfect, but I can't remember any reviews where that was the case across all titles and often the difference comes from the 0.1% lows.
 
With the updates over the past year (or so?) to the Xbox Game Bar and the AMD Chipset Drivers, has your source re-run those tests?

Honestly I think you just hit the nail on the head as to why the 7800x3D is the better gaming CPU. Having hardware with fundamental issues that need to be baby-sat with software solutions at best, or straight-up manually adjusted (process laso, etc) at worse.

Constantly having to hope that the chipset drivers have what you need, that the thread scheduler and/or Xbox game-bar gets everything right, that new games work perfect with everything at launch, and always praying that new updates to all of these things come out that might fix problems that should have never existed in the first place.

Or just get a single-CCD chip like the 7800x3D and literally never have to worry about any of that, ever.

It reminds me a lot of SLI/Crossfire and constantly hoping that each game had a good working SLI/Crossfire profile because you were completely dependent on that.

And just like how SLI/Crossfire setups stopped working with new games once they stopped releasing new SLI/Crossfire profiles, what makes you think that the same thing won't eventually happen with these CPUs? How long into the future do you think that AMD is going to continue to supply updates to make sure that new games end up on the correct CCD? At some point, they are just going to stop and only focus on their newer CPUs, at which point it will become a shit-show which CCD games end up on. Potentially I could see something similar happening with Intel chips and games ending up on E-cores, but somehow Intel chips are not dependent on the XBox game bar whereas AMD chips are.
tard.gif
 
Honestly I think you just hit the nail on the head as to why the 7800x3D is the better gaming CPU. Having hardware with fundamental issues that need to be baby-sat with software solutions at best, or straight-up manually adjusted (process laso, etc) at worse.

Constantly having to hope that the chipset drivers have what you need, that the thread scheduler and/or Xbox game-bar gets everything right, that new games work perfect with everything at launch, and always praying that new updates to all of these things come out that might fix problems that should have never existed in the first place.

Or just get a single-CCD chip like the 7800x3D and literally never have to worry about any of that, ever.

It reminds me a lot of SLI/Crossfire and constantly hoping that each game had a good working SLI/Crossfire profile because you were completely dependent on that.

And just like how SLI/Crossfire setups stopped working with new games once they stopped releasing new SLI/Crossfire profiles, what makes you think that the same thing won't eventually happen with these CPUs? How long into the future do you think that AMD is going to continue to supply updates to make sure that new games end up on the correct CCD? At some point, they are just going to stop and only focus on their newer CPUs, at which point it will become a shit-show which CCD games end up on. Potentially I could see something similar happening with Intel chips and games ending up on E-cores, but somehow Intel chips are not dependent on the XBox game bar whereas AMD chips are. View attachment 640845
Damn, didn't realize every single GPU, motherboard and most usb devices I've ever used are fundamentally flawed and need to be baby-sat with software solutions, too. We should all go back to CPU rendering on only hardware that has native support in windows because the .1% lows were bad in a review a year ago in a 12 year old game. Guess what, if you didn't have software support for your 7800x3d core priority and a modern scheduler, it'd suck too. It would not prioritize the highest boosting cores first in low core loads and wouldn't spread the load across cores to reduce localized heat further reducing boost. It'd be worse without support, but I guess that's ok because there's a driver that ships with windows for that instead of being part of the chipset drivers (or is it)? Sure, that makes sense.

Except every single game I run now has no micro stutters or issues and the only thing I had to do was install the motherboard and chipset drivers and it just worked perfectly. Exactly the same thing I'd have done with a 7800x3d. Oh, and those have micro stutters, too, an internet search says so. What's the problem there?

Listen, the 7950x3d is dumb if you're just gaming, I don't think anybody is arguing that. But it's the same as a 7800x3d to slightly better, so there are use cases for it. Because you don't have the skill to get your system working properly doesn't mean it's not cake for others.
 
Back
Top