[ryouiki]

Reported Issues Status

Issue | | Status
Disabling CSM causes BIOS to become slow||Reproducible / Working on a fix*

Is this still being worked on? I just installed a X570 Aorus Master and have to disable CSM... the severe lag in the BIOS is about the only complaint I have with this board, it makes doing things like inputting all of the memory timings/subtimings extremely annoying.

 

[ryouiki]

Oh wow! I haven't been able to see any core above 4550ghz since 1003ABBA. Increasing BCLK clock might help but not sure if I want to mess with that right now.

I have seen single core boosts reported up to 4650 on F11 without adjusting BCLK... however this only occurs if VCORE is set to normal/0 offset. Applying any type of negative offset eventually results in max reported boost slowly getting lower. Below 4600 by -0.0185, 4525 by -0.050, etc.

That said, single core benchmark scores are roughly the same even though the reported value for boost is less.

 

[ryouiki]

Can someone else test the svm performance decrease?
Im sure svm on kills bclk values stability. I mean system works fine, just no 100mhz bclk even if you set higher manually.

This decreases max clocks throughout, affecting cores, ram, gpu even nvme scores. About 2% penalty using hyper-v for gigabyte boards

I turned on SVM, powered off/restarted and did a few tests on my Master w/ F11. Bus clock still reporting fixed at 100.0, Cinebench R20 run actually scored slightly higher then the previous run, but that is probably run to run variance due to this chips heat sensitivity.

Hard to say on max clocks, I only rarely see 4600 on the best core, but that could take anywhere from minutes to hours to be seen.

 

[ryouiki]

My Aurus Master incorrectly sets the RAM voltage. instead of 1.35 it shows 1.38. When I left the settings on the auto, the same thing happened. I changed to manual settings, as you see the same thing.

There is a post somewhere that talks about this (which I cannot find at the moment)... Gigabyte's response was something along the lines of them doing internal testing and finding a slight deviation between set voltage, reported voltage, and actual measured voltage. If I remember correctly there was something like 0.009 variance they found.

*Edit* believe it is here:

https://www.overclock.net/forum/11-...570-aorus-owners-thread-101.html#post28078624

 

[ryouiki]

Anyone have any ideas on WHEA errors I am getting:

Event 19 - WHEA Logger:

A corrected hardware error has occurred.

Reported by component: Processor Core
Error Source: Corrected Machine Check
Error Type: Bus/Interconnect Error
Processor APIC ID: 0

I can't quite figure out what is causing these. They only seem to happen when the system goes from actual utilization into idle state... I can for example run stress tests/benchmarks/games for days without seeing these, and then once I exit and system goes into idle they will show up occasionally (maybe one every few days).

My first thought was this was related to SOC voltage/running 1900 IF/MCLK.

 

[ryouiki]

I must agree - sound like IF.

I read something about that raising VDDP/VDDG and Vsoc also helps stabilize.
But anyhow IF at 1900 is at the limit - not all Ryzens will do it stable.

Greetings,
Medizinmann

I am going to try to raise VSOC a little more and see if that has any impact... the fact that it only every occurs during idle/lower power state makes me feel like some voltage is maybe not stable, otherwise there is no other issues (no crashes/lockups/10000%+ stable on memory testing etc.).

Power Down on memory is disabled. The only other posts I could find about similiar WHEA (bus/interconnect) errors were occuring on systems with Northbridge and raising Northbridge voltage fixed their issues.

 

[ryouiki]

Anyone have any ideas on WHEA errors I am getting:

Event 19 - WHEA Logger:

A corrected hardware error has occurred.

Reported by component: Processor Core
Error Source: Corrected Machine Check
Error Type: Bus/Interconnect Error
Processor APIC ID: 0

If someone else ends up running into this, I appears raising VSOC appears to resolve this... changed the BIOS value slightly (1.112 I believe) and haven't see this again.

 

[ryouiki]

Anyone else get an error trying to install this?

There were some people complaining about this in a thread on another website, but I didn't have any issues. The only update in this release compared to the previous one is the PCI driver though, everything else is the same.

 

[ryouiki]

Really depends on your case. My case lacks a side intake fan due to glass panel so my chipset was always 58 before mod even though I do not have gpu dumping hot air into the pch intake.

EDIT: Just note 60 is the point where it will kick on if using the silent profile

Maybe a stupid question but what is the difference between the readout from the X570 Chipset section and the IT8688E Chipset section? The former can reach 70C on my board, the later usually stays below 50.

 

[ryouiki]

Cold boot is pretty slow, but soft reboot seems more normal.

The one thing I find so odd though is when just now I moved my Samsung nvme drive from M2A to M2B socket, and then turned on the PC, it got stuck at Od hex code, and then eventually turned itself off. Had to turn system on again and then bios was reset, forcing me to go back and redo all my settings.

At least I have the nvme thing fixed, turned out I just needed to install the windows driver for it (so not related to socket).

Bios has been reset maybe 3-4 times now since last night. This is a real pain in the butt.

Cold Boot will be slow, my Crosshair is probably slightly longer to boot then the Master.

As for the "BIOS reset" is it possible it is failing boot and switching to backup BIOS? I personally flip the dip switch to force the board into single BIOS mode when first setting up/changing memory timings etc... otherwise if you have a issue on boot you can end up in backup BIOS which will then have completely different settings.

 

[ryouiki]

Man, this Gigabyte bios is still weird. I changed the dip switch to manual bios selection. Turn back on the pc, takes forever to do something. Then after a minute or two I finally see the post screen, so hit Del, and to my surprise it had reverted back to the F4 bios, even though I had flashed F11 right when I first got it on Tuesday.

I just can't shake this bad feeling I have about this board, just not leaving me with a good impression. I'll have to take a closer look at those Asus boards, though the only one with 2 internal USB 3 headers is the C8F, at $560. Ugh.

There are two BIOS chips on the board (Main/Backup)... if you have flipped the dip switch to manual and end up in a different BIOS version then you are booting off the other chip (which one I can't be sure). If the dip switch settings are set to boot off main bios, then you probably had stability issues forcing you into the backup chip which you flashed/changed your settings on. If you aren't sure which chip it is booting from there are two LEDS to indicate which chip, however these are not so easy to see if you have a large video card in the first PCI-E slot.

This is why I don't like automatic switching of BIOS chips by default... I will flash the main bios, set the system up and make sure it is 100% stable, and only then manually force to the backup and repeat the process, loading the profile dumped from the main bios.

 

[ryouiki]

My question thusly is: what do other folks see on a 3900x in CB in terms of frequency? I wonder if something is not done right with the cooler setup that the temp shoots up so quickly.
I have tried the same benchmark with the case open and the AIO fans blowing at 100% with no change.

CB20 has my 3900x at about 4.10 - 4.15 and peaks at about 72C under a NH-D15 and scores around 7350. That is pretty much stock settings + IF1900/3800CL16, and a negative voltage offset of 0.050 or so.

*Edit correction to IF clock*

 

[ryouiki]

Hey guys, I think I need some help tuning my memory with my Ryzen 9 3900X. No matter what I do I just can’t get these sticks to work with any config beyond stock. I believe the IMC on my 3900X is total garbage. It’s a little bit disappointing to see since I bought one the best binned kits supposedly. The G Skill Trident Z 3600MHzCL15 kit 16GB (2x8GB) F4-3600C15D-16GTZ.

I have two kits of this memory, and was never able to get stability beyond tightening timings @ 3600. This both on Aorus Master and a Crosshair VII Hero... I thought I had a terrible IMC too, replaced this kit with 4x FlareX 3200CL14, and immediately was stable at 3800Cl16/IF1900.

I don't know what it is about this particular RAM kit, but neither of my AMD motherboards likes it.

 

[ryouiki]

That's interesting to hear, I thought that since this kit was binned to this extent it would play nicely with Ryzen. Had I known I wouldn't have paid a premium for it.

Somewhere in the Crosshair VII hero thread there is a post showing someone running this kit at 3800 on a 3800X I believe, but I was never able to get the same results with any stability (I couldn' even successfully boot in that configuration on the C7H) on either kit I own (they are sequential serial numbers).

I know when I asked in that thread at least one person stated that they start acting funny above 1.43V.

I had none of these problems with the FlareX kit even with 4 DIMMS.

 

[ryouiki]

BIOS F12E Aorus Master Quick Test:

1.) Still lagging with CSM disabled

Does the workaround from F7B (CTRL+F6) exist in this version at all? It isn't present in F11.

 

[ryouiki]

Sorry. This is the first time I heard about that. So no Idea what you are talking about. Also I am currently running HCI Memtest - can not play with the BIOS for a good while.

I think it is mentioned on the 3rd or 4th post in this thread... supposedly F7B had a workaround to the slow CSM by allowing you to press CTRL + F6 and switching to "VGA First" which lowers the resolution of the BIOS but removes the delay.

 

[ryouiki]

not even reddit knows.
https://old.reddit.com/r/Amd/comments/flibr9/amd_chipset_drivers_31920_version_203120657/

comment suggest there were only 2 changes:
https://old.reddit.com/r/Amd/comments/flibr9/amd_chipset_drivers_31920_version_203120657/fl0mzsb/

The PCI Driver and GPIO driver were updated, everything else is the same.

If you dig into the extraction directory (C:\AMD) there are actual release notes for the GPIO driver:

The EOI bit is set for bank based on the number of banks present. This change fixed system hang issue on Stoney based systems.

The PCI driver has no information.

 

[ryouiki]

I guess the FCLK on my 3900x is just on the borderline of stability. I posted previously how I would occasionally get WHEA errors about "Corrected Machine Error - Bus/Interconnect", and I thought adjusting VSOC had solved this but they are showing up again.

The oddest bit is that I cannot get these to occur under load, only when the system is idle. Sometimes I have to leave the system sitting idle 10+ hours to even see one, sometimes they occur in 1-2 hours.

Adjusting FCLK/MCLK to 1867/1867 has left the logs without any errors for about 3 days now... I am not sure if there are any voltages other then SOC/VDDP/VDDG I can try to play with, or if this chip is just not 100% capable of sustaining 1900 FCLK without error. I've already attempted to push those about as high as I am comfortable with but the problem persists.

Even when it does error, it doesn't impact anything as apparently since it is "Corrected", but I don't consider it stable that way. Anyone have any other ideas?

 

[ryouiki]

Some one posted in one of the other forums that bumping up the chipset vcore can help with 3800/ 1900 IF stability.

It isn't clear to me, these are the PM_* values in the BIOS?

Also which BIOS versions did you tried with?
So far the most stable and "bug free" for me it's F12a.

What about the SOC LLC setting? Did you try with Low or Medium while lowering the voltage?
With lower SOC voltage you can also try lowering the VDDP/VDDG, it can improve reliability for IF and RAM.

I've only every used F11, still waiting for an "official" F12 release. I tried a number of various settings for SOC (Fixed 1.1, Offset, etc.) + SOC LLC without much luck, though this is incredibly difficult to test because it can take so long for one of these errors to show up in the Windows event log. The only thing I can say for sure is lowering VDDG below about 1000mv @ 1800 FCLK will result in actual crashing.

As far as the memory it is using custom timings (baseline from DRAM calculator)... which has passed multiple runs of Karhu running 12+ hours or more... since memory testing keeps the system at least lightly loaded the WHEA errors have never appeared while it is running.

 

[ryouiki]

I'd give a try to F12a; I used F11 for quite a lot and at least for me it's more reliable.

... snip ...

Best way to test stability is gaming with some medium to high workload; something too heavy will not trigger that easy instability.

I'll probably wait until F12 official for now. I tried raising PM_* voltages but no change, WHEA error will still come after some undetermined amount of time. Strangely I did notice that power plan seems to impact this... if I use a more aggressive plan like 1usmus, then it seems much less likely to happen.

My suspicion is that there is some power saving going on when system is idle that triggers WHEA error... the one thing I haven't tried is the option in the CBS -> XFR menu that disables the SOC from doing any power saving... I may try that later this weekend.

Otherwise RAM tests all pass, I can run any type of benchmark/games for 10+ hours and never see an issue... as long as I don't let system sit in Windows idle for extended amount of time. Ex. yesterday I played Doom Eternal for like 5-6 hours... exit game, go to other room and come back and WHEA error has triggered while sitting at Desktop.

None of this occurs if I just lower FCLK to 1867.