AMD Ryzen 7 5800X3D

[gffermari]

-The background apps affect the results. I closed most of them and had an increase in cpuz and R23.
Also in the AI Suite I managed to increase the cpu power. From 90 to 124.5 watts during multicore benchmarking in Cinebench.
Note: In R23 now I got 86 degrees after 4-5 loops and I'm on three 240 radiators....
But the cpu doesn't throttle while the water temp increased by just 1.5 degrees.

The thing is that I'm still behind in numbers.

Regarding the Windows thing. I have to admit that I'm in the worst possible position.
I have Windows 11 for almost a year now, it was upgraded from Windows 10 which was installed about 3-4 years ago...

CPUZ: 617/6180

Guru3D	639	6599
Toms Hardware	627	6070

Most reviewers do not use cpuz. Googling you can see screenshots of 5800x3d cpuz numbers which are significantly higher than mine.
example:

Cinebench R23: 14175 (I haven't tried the single core benchmark)

TechpowerUp	1496	14688
Techspot	1398	14221
techradar	1483	15070
PCWorld	1490	15014
kitguru	1495	15171
hothardware	1486	14612
overclock3d	1452	14461
computerbase	1488	14907
	1479	15031
	1484	14739

 

[529th]

@gffermari​

CPUz ST bench mark is on par at 617 if you have Global C-States disabled .. possibly some other bios tweaks in there causing it also. What's it set at?

Those other results are from bus overclocking

 

[R0H1T]

Should set CPU Power & VDDSOC phase controls to optimized & turn off spread spectrum under CPU VRM Switching frequency. This wouldn't make a big difference but perhaps enough in your case.

 

[Mussels]

CPU-Z's stock results are always higher than mine

There was also bugs with windows 11 where if you changed processor you had to reinstall windows clean to fix performance issues - i never heard back from people i suggested try an 'upgrade' install if it helped

 

[Count von Schwalbe]

There was also bugs with windows 11 where if you changed processor you had to reinstall windows clean to fix performance issues - i never heard back from people i suggested try an 'upgrade' install if it helped

Did the Ryzen L3 bug on W11 get a final fix or just a workaround? As the L3 is the reason for the existence that could be an issue...

@gffermari have you tried clean installs of W11 and W10? Just a thought

 

[gffermari]

Thank you all. I tried every possible setting combination in the bios but in most cases I get a lot worse numbers. (Meaning CPUz: 590/5800 range)
I will install windows in a second ssd and check.

 

[Taraquin]

I hope AMD makes curve optimizer available in future agesas, the 5800X3D sure needs it. Until the the pbo2 tuner tool can be quite helpful and lower temps a lot and even boost performance is cooling is not vety good.

 

[Mussels]

Did the Ryzen L3 bug on W11 get a final fix or just a workaround? As the L3 is the reason for the existence that could be an issue...

@gffermari have you tried clean installs of W11 and W10? Just a thought

the L3 bug got its fixes, this CPU upgrade one seems to be seperate - people kinda stopped talking about it

 

[Amaze]

New bios update on the way:

AMD AGESA V2 1.2.0.7 Microcode To Fix fTPM Stutters

Motherboard manufacturers are addressing AMD's stuttering issue in Windows 10 and 11 with a new BIOS update. The (intermittent) performance losses and stuttering are caused mostly by the TPM 2.0 func...

www.guru3d.com

 

[Chomiq]

New bios update on the way:

AMD AGESA V2 1.2.0.7 Microcode To Fix fTPM Stutters

Motherboard manufacturers are addressing AMD's stuttering issue in Windows 10 and 11 with a new BIOS update. The (intermittent) performance losses and stuttering are caused mostly by the TPM 2.0 func...

www.guru3d.com

There's already news about this on TPU and 1.2.0.7 bioses were out a month ago.

 

[chrcoluk]

How many microcodes do they need to fix fTPM stuttering? I count 3 so far. Will 1.2.0.8 fix it as well?

 

[dont whant to set it"']

@chrcoluk are you up to something?
I've been on the "1.2.0.6c beta" with my 5800x3d.

 

[Amaze]

There's already news about this on TPU and 1.2.0.7 bioses were out a month ago.

It says V2.

Edit:
It is indeed old. My board already has this as well. Who knows why this is news then.

 

[Deleted member 24505]

How many microcodes do they need to fix fTPM stuttering? I count 3 so far. Will 1.2.0.8 fix it as well?

They're getting as bad as Intel now............oh wait Intel just works.

 

[Count von Schwalbe]

They're getting as bad as Intel now............oh wait Intel just works.

Not illogical as Intel has dominated the market for years, well-nigh forcing MS to develop around their hardware. However, this is strongly off topic.

 

[TheoneandonlyMrK]

They're getting as bad as Intel now............oh wait Intel just works.

It's like spectre didn't exist in your head, wait long enough and spectre MK2 will release, passing you an Intel microcode update that fixes stuff by tanking performance.



Now on topic I have not had these stutters with my RYZEN system despite being on windows 11 and using the allegedly enhanced security fTPM enabled etc offers, odd.

 

[Deleted member 24505]

Funny how it's different when it's a Intel thread though with off topic anti Intel posts, usual TPU double standards now though.

 

[TheoneandonlyMrK]

Personally I'm fine with your comments, they need countering(in any thread where I disagree with you I'm there) though but troll on in this or any Intel thread it's a public forum after all, but man do you moan.

 

[Tomorrow]

Funny how it's different when it's a Intel thread though with off topic anti Intel posts, usual TPU double standards now though.

^ Says the resident Intel fanboy. If you dont have anything meaningful to add to this topic then dont post here.

 

[Mussels]

How many microcodes do they need to fix fTPM stuttering? I count 3 so far. Will 1.2.0.8 fix it as well?

1.2.0.7 was the only one to ever fix the stuttering, since it was caused by 1.2.0.6 (which added the windows 11 support)


It's not like we could just disable fTPM and have it all work fine, the entire time...

 

[Deleted member 24505]

If you dont have anything meaningful to add to this topic then dont post here.

The same should apply to Intel threads then.

 

[Tomorrow]

The same should apply to Intel threads then.

Yes it should. But someone else doing it in Intel threads does not excuse you doing it in AMD threads.

 

[chrcoluk]

@chrcoluk are you up to something?
I've been on the "1.2.0.6c beta" with my 5800x3d.

I am on 1.2.0.6 as well. Working great for me.

 

[gffermari]

Fresh Windows install today.

Results:
CPUZ ST = 610-615 the mediocre cores, 621 the best core.
CPUZ MT = ~6225. I get random results every time and it doesn't follow a pattern. No matter what settings I set, the score is from 5800 up to 6225(best) while in every test it begins at 6300 and gradually decreased to a random number....5900, 6150, 6025, whatever....

R23 MT = ~4250. No matter what settings I put on BIOS, I can't get over 4200 points.

OCCT ST (SSE, AVX) = I get the lowest score for a 5800X3D - slightly behind or in front of the lowest 5800X3D score in the benchmark.
OCCT MT (SSE, AVX) = I get slightly better score than the average score of a 5800X3D.

I don't know if my RAM causes all these issues but just today I noticed that I get some weird results in OCCT.

----------------------------------------------------------------------
edit....

....at last, i managed to unleash some of its power!!!

 

[Amaze]

Fresh Windows install today.

Results:

R23 MT = ~4250. No matter what settings I put on BIOS, I can't get over 4200 points.

That is extremely low. The best I could do so far with the PBO curve tool was 15013.

How-to-undervolt-AMD-RYZEN-5800X3D-Guide-with-PBO2-Tuner/README.md at main · PrimeO7/How-to-undervolt-AMD-RYZEN-5800X3D-Guide-with-PBO2-Tuner

Get the Most out of your 5800X3D using PBO Curve Optimizer! - PrimeO7/How-to-undervolt-AMD-RYZEN-5800X3D-Guide-with-PBO2-Tuner

github.com