AMD Zen 2 (Ryzen 3000) - *** NO COMPETITOR HINTING ***

[fs123]

Spanish article here:
https://www.xanxogaming.com/reviews...s_WHEA_PCI_Express_en_tarjetas_NVIDIA_GeForce

Comments from le Reddit:

"Driver level errors associated with either the PCI Express bus and/or Nvidia GPUs may have prevented the chips from boosting properly - which was interesting because several reviewers today commented on not hitting advertised boosts. The author observed this behavior on both Gigabyte and MSI boards but only with Nvidia cards and at least for their own work decided it was sufficient enough to warrant not publishing any FPS benchmarks until sorted out."

"This is huge dude, he said he got boost to 4.65 GHz with the older 1.0.0.2 bios, even though PBO did not work on that BIOS. The newer 1.0.0.2 (NPRP press code) & 1.0.0.3AB BIOS were limiting his boost to 4.35 GHz
The driver WHEA errors seem to be unrelated and are a nVidia specific problem. They are probably still causing slowdown in benchmarks."

Thread:
https://www.reddit.com/r/Amd/comments/cacasv/very_interesting_writeup_of_an_issue_that_may/

Here's the website translated to English..
https://www.xanxogaming.com/reviews...EA_PCI_Express_errors_on_NVIDIA_GeForce_cards

This is what they wrote about the issue..

During the first three hours of testing of the AMD Ryzen 9 3900X processor, using the X570 AORUS XTREME board, I noticed the problem when PCMark 8 did not pass the first test after 40 minutes (this is a total of ten tests). I noticed WHEA error (Windows Hardware Error Architecture) in HWInfo64 (se this software for PC telemetry, highly suggested).

From there I also decided to pay more attention to HWInfo64 and also checked that the BOOST frequencies of the processor had problems, since it didn’t get to “boost” all its cores to the maximum that it should, which is 4.6 GHz. It reached 4.5 GHz to 4.575 GHz in a pair of cores and the rest of cores to 4.3-4.4 GHz… We used manufacturers chipset driver, we have used press chipsets, as more current chipset driver version, same results.

I thought boosting on 1 or 2 cores was normal. Is it really supposed to boost all cores?

 

[Dirk Diggler]

Tiny Tom said something along the lines of:

For anything 3700x and below you are OK with a x470 and below board. For anything 3900x you will want to get a x570 mobo for the vrms etc...

Your mileage may vary.

That's a bit concerning, I was under the impression it was only the 3950x that would require x570.

 

[TrM]

That's a bit concerning, I was under the impression it was only the 3950x that would require x570.

I doubt any X470 motherboards will struggle with 3900x or even the 3950x I feel that it’s just another upgrade con myself and people are believing u need 12 14 heck even 16 power phases to get a stable overclock. Maybe if we could push the frequency way up but with the 4.6 realms I think any X470 mobo will do

 

[Sin_Chase]

Whilst this speculation about a PBO boost issue is interesting I'm not sure it's as big as reported.

Whilst not boosting correctly would be an issue reviewers are still not getting stable overclocks with the manual offset overclocking. It's also readily apparent there is a thermal ceiling when trying to dial in any more voltage.

Interesting nonetheless though, especially if 4.6 was being achieved on earlier BIOS revisions.

 

[BoomAM]

Does anyone know what the default voltage of a 3900x is meant to be?
For the life of me i cant find it documented anywhere.

 

[Aretak]

That's a bit concerning, I was under the impression it was only the 3950x that would require x570.

It isn't true even for the 3950X. My Crosshair VI Hero, an X370 board, has more than enough VRM to handle the 3950X. And that's not just a guess or a prediction - it's knowing exactly which components the VRM uses (TI CSD87350Q5D NexFETs), exactly what current output they can provide (40A) and the total current output that the VRM can provide (320A). Which is way, way more than even a 3950X is ever going to pull. Current capability is ultimately more important than the number of phases a VRM has, which is why Asus have hugely reduced the number of phases on a lot of their newer boards, but bulked up each phase by using double the number of components, increasing the current capability of the VRM without needing an overkill phase count. It's still a cost-saving measure on their part, but it's ultimately not one that makes their VRMs "bad" or incapable of handling the thirstiest CPUs. Just at the expense of potentially more voltage ripple than a higher phase count VRM, though how much that matters in real world terms is debatable (you still see people pushing 9900Ks to the limit on the Maximus XI Hero for example, which is a mere four-phase VRM).

Aside from all that, despite how capable the Crosshair VI Hero's VRM is, most high end X470 boards are even better, so will also easily handle the 3950X. Other high end X370 boards might prove to be a trap to people though, as MSI and Gigabyte's efforts were pitiful, with weak VRMs. The X370 Taichi should also be more than capable of handling a 3950X though, along with the Crosshair VI Extreme.

 

[Colonel D]

Is 3700x still available or have the pre orders all been allocated?


Would like to go for 3700x and the x570 gaming edge

 

[Panos]

Whilst this speculation about a PBO boost issue is interesting I'm not sure it's as big as reported.

Whilst not boosting correctly would be an issue reviewers are still not getting stable overclocks with the manual offset overclocking. It's also readily apparent there is a thermal ceiling when trying to dial in any more voltage.

Interesting nonetheless though, especially if 4.6 was being achieved on earlier BIOS revisions.

Or AMD cards apparently, as the article states the errors do not exist with RVII or V56.

 

[StriderX]

I wonder if because the hardware was quicker to production than they thought it would be, that the software side has been rushed to "just working, get it out now".

 

[BoomAM]

Same as every other launch.
All doom and gloom from people going 'pfft, typical AMD' even though ALL the manufacturers do it, in all industries.

My concern right now is temperatures and voltages.
Voltages seem to bounce quite often up-to 1.5v (and thus cause high idling temps), and for the life of me i dont know if that's normal for the 3900x or not.

 

[Panos]

I wonder if because the hardware was quicker to production than they thought it would be, that the software side has been rushed to "just working, get it out now".

Apparently the older versions of the BIOS do work without issue. The new one has the issue.
Tbh isn't that big thing. AMD can sort these things out within the next 48 hours. Except if it is truly NVIDIA drivers issue. That should rely on NV to resolve.

 

[Panos]

Same as every other launch.
All doom and gloom from people going 'pfft, typical AMD' even though ALL the manufacturers do it, in all industries.

My concern right now is temperatures and voltages.
Voltages seem to bounce quite often up-to 1.5v (and thus cause high idling temps), and for the life of me i dont know if that's normal for the 3900x or not.

PBO can push the voltages to some high values even on 2000 series. This is done briefly and within the parameters, assuming you are using PBO and not manually set the voltage.

 

[BoomAM]

PBO can push the voltages to some high values even on 2000 series. This is done briefly and within the parameters, assuming you are using PBO and not manually set the voltage.

I've not, but i have been trying to get idle temps under control.
Even with a manual offset of -0.150 the chip seems to idle in the mid-40's, or low-50's when kept stock. Is that normal for a 3900X on a AIO (H100i v2) ?
Had a play around here if it helps for context.

 

[Panos]

I've not, but i have been trying to get idle temps under control.
Even with a manual offset of -0.150 the chip seems to idle in the mid-40's, or low-50's when kept stock. Is that normal for a 3900X on a AIO (H100i v2) ?
Had a play around here if it helps for context.

Temps for a 12 core 24 thread chip aren't that bad as other said. I checked atm my 8600K @ 4.5Ghz, and has 40C idle temps with a Noctua NH D15 at low rpm.
Make sure you have fit it correctly. Also have you tried to let the CPU run without fiddling with PBO etc?

 

[BoomAM]

Temps for a 12 core 24 thread chip aren't that bad as other said. I checked atm my 8600K @ 4.5Ghz, and has 40C idle temps with a Noctua NH D15 at low rpm.
Make sure you have fit it correctly. Also have you tried to let the CPU run without fiddling with PBO etc?

Yes.
One thing i have just noticed, the Corsair iCue software linked my fan curve to the water temp for some reason. Linked it to CPU temp and now after 10mins or so of Prime95 it doesnt go above 65*c load.
Idle is still higher than i'd like, but i suppose for a chip thats uses a good 40-50W more power and has 50% more cores than my older 2700 its not too bad to be in the high 40s, low 50s range.

re: fitting, yes.
I use my age old technique of applying a thin layer across the whole head-spreader, and then a small blob in the center to account for any inconsistencies in the middle of the spreader or cooling element.

 

[Panos]

@BoomAM I found this on this video

https://www.youtube.com/watch?v=WXbCdGENp5I&t=10s

Which BIOS revision did you use? Because The Stilt has this to say:

"In case of the ASUS Crosshair VIII Hero Wi-Fi motherboard, the media was instructed to use 0066 bios build, which had been vetted and approved by AMD.

However, newer bios builds were available and ASUS has also (allegedly) told the media to use those versions. What exactly has transpired here is still under investigation, but regardless of the actual reasons behind it, the consequences might be rather significant.

In practical terms, all reviews which were done on ASUS Crosshair VIII Formula or Hero motherboards using other than 0066 bios build must be considered invalid, at least partially. Reviews using other ASUS motherboard models (not provided by AMD) are under suspicion as well.

Few days ago, I noticed certain anomalies, while measuring the power consumption of the different Matisse SKUs. Inspection of the power management parameters revealed no issues, which could have explained those anomalies.
The external power measurements (VRM DCR) revealed that the CPU was consuming significantly more power, than its power management should have allowed it to.

I initially suspected that this was AMDs own doing, in an effort trying to boost the performance of the new CPUs even further, but further investigation indicated otherwise. AMD had no part in it, and the actions by ASUS are the sole reason behind it.

The investigation revealed that ASUS is altering one or more power management parameters of the CPU, causing it believe it consumes less power than it actually does. As a result, the frequencies will be higher than the actual power budget would normally allow to. Tricks like this are pretty much a common (mal)practice these days however, there is a good reason why this must be considered worse than the others: this "thing" is completely undetectable without external measurements and rather deep knowledge, but also there is no way to disable it either. Even a person such as myself, who can control most things on these platforms cannot disable this "thing".
As you may notice, at the moment I call this issue the "thing", since I'm giving ASUS the benefit of a doubt."

Because is 3:30am here, and had few drinks tonight, I do not remember your board, but if you have ASUS ones, it might be their usual overzealous approach on power delivery

 

[humbug]

So in some reviews they are not quite boosting as high as they should because of some BIOS /Driver bug, is that it?

 

[Panos]

So in some reviews they are not quite boosting as high as they should because of some BIOS /Driver bug, is that it?

According to the big spanish article, the boost clock issue occurs with the PCIe when using Nvidia GPU only.

The CPUs do not boost properly, and there are PCIe errors. However using older BIOS resolves the issues, or by using AMD GPU.

 

[humbug]

According to the big spanish article, the boost clock issue occurs with the PCIe when using Nvidia GPU only.

The CPUs do not boost properly, and there are PCIe errors. However using older BIOS resolves the issues.

So when using nVidia "It just doesn't work" yeah ok Jens...

 

[ltron]

I'm going to wait for the dust to settle before ordering, supposedly there are AGESA bugs affecting gaming performance and boosting behaviour. Hopefully these problems get fixed soon: https://www.reddit.com/r/Amd/comments/cacwf9/psa_ryzen_3000_gaming_performance_is_being_gimped/

I hope someone will do a comparison of X570 chipset fans and which are controllable as some early user reports suggest some are better than others, take it with a grain of salt until this can be confirmed by reviewers.