[Sphex_]

Matthew, we need a time-based hysteresis for fans, you can check how Asrock does things.

+1 on this. Ryzen's temperature jumps around way too much for a temperature based hysteresis to be effective. ASUS also has this feature, except it never worked on my X470 motherboard lmao.

 

[Sphex_]

Hey. So I just updated my AORUS x570 ELITE bios to the latest version F4j from F3. I re-applied the XMP profile and put precision boost overdrive back on enable like before as the update reset everything to default. I've just tested cinebench R15 and before the bios update my score was 1583 now after the update it has dropped to 1409 with a Ryzen 3600. Anyone know why it may have dropped nearly 200 point because of the bios update?

Edit: I seem to of lost 165Mhz with the precision boost after the update. Before it was boosting to 4165MHz now it's boost to around 4000Mhz. Anyone know why and how to fix it?

Unfortunately PB basically got nerfed in the later BIOS revisions due to an update in AGESA code. F3 is AGESA 1003, whereas any release afterwards is 1003A+. For whatever reason PB clocks tanked and PBO is just straight up broken. The only fix is to either flash back to F3 or wait for an update from AMD and Gigabyte. :/

 

[Sphex_]

Yip, RGB Fusion is really bad..... Even when it's not triggering Anti Cheat software.



I let the software auto start and have a Scheduled Task which runs at login with elevated privileges which runs a batch script. The script waits 20 seconds for the software to load and set colours and then terminates CheckKill and RGB Fusion processes.

Huh. I just unchecked the "Startup with Windows" box or whatever it is in RGB Fusion. 0 Problems with anti-cheat or anything else, and the motherboard retains the color I've set through shutdowns and startups. Only Process I see right not associated with Gigabyte is "GCloud / GService".

 

[Sphex_]

Have you tried with Gear down mode off? It usually adds a bit of latency. Also turn off power down mode. Also, tRFC2 & tRFC4 should be lower than tRFC.

Ryzen doesn't look at tRFC2 and tRFC4 values; they have no effect.

 

[Sphex_]

PBO is broken for all versions of motherboards?
I am just trying to understand if there is any difference between PBO in Auto and Enabled mode at my Elite.

Sent from my ONEPLUS A6003 using Tapatalk

I flashed F5a to my Aorus Elite and PBO now kind of works. I'm not really sure if it does anything when set to Auto, but when setting it to "Enabled" in both the AMD CBS and AMD Overclocking menus, the power and current limits (PPT, EDC, TDC) are raised to crazy values and I definitely see an increase in sustained all-core boost clocks. It's also reflected in a ~200 point uplift in CB20 multithreaded scores. Setting either of the PBO values to Manual or Advanced results in VCore being capped to like 1.44V, crushing performance.

 

[Sphex_]

How does your CB single score fare with PBO on?

On my setup it drops because the core clock will not go above the 4.150ish that I get with multi core cb15 run which does go up

curios to see if you have better results with single core and PBO on

edit: on my setup with PBO on the vcore is capped @ 1.39 even on single threaded loads. With PBO off(auto) the vcore is now capped @ 1.5 definitely a bug

Well in Cinebench 20 I was scoring around 4800 Points during the multithreaded test with PBO set to Auto, and scored 4967 when set to Enabled. I don't think I ran CB20 with single thread before but I see you're running CB15 so the results aren't really comparable anyways. All I know is I got a higher score and core voltages looked much better than what they did when set to Manual or Advanced. I'd say setting the PBO to Enabled is basically the same as having it set to Auto except now the thermal and current limits are raised so the processor doesn't throttle itself so soon. I still see 4.15 GHz All-core during a CB20 MT run and the CPU gets ~80°C even with a stout air cooler mounted to it.



I agree though, I don't think I've ever seen my VCore exceed 1.5V no matter what my PBO settings are, that might just be a cap set by Gigabyte or AMD though.

 

[Sphex_]

Why are you guys running dram voltage so low i saw 1.37 i run mine at 1.5 no issue no instability not really hot either.

Some kits become unstable with too much voltage. For example, my particular B-Die kit will fail Memory stability tests at 1.44V but will be stable at 1.43V. More isn't always better / safer. Some kits, depending on the dies, can't run 1.5V. Not every kit is like B-Die. In reality, you should run the lowest voltage (or very near the lowest voltage) that your memory kit can remain stable. Micron E-Die in particular seems to be able to achieve relatively high memory speeds at impressively low voltages, especially when compared to other memory die types.

 

[Sphex_]

Came across this in reddit and though maybe those of you here with issues can try to see if this help you

With the new bios update for gigabyte x570 mobos Agesa 1.0.0.4 , many people started experiencing performance loss for cpu boost clocks. While some gained higher cinebench scores but lower clocks, others like me have experienced a total loss in both clock and cinebench score.

The problem was no matter how pbo was adjusted in the bios cpu voltage remained at a constant 1.2v-1.25v, hence cpu clock speed never increased. This was happening because the EDC Limit is scaling backwards now. I don't know if this is intentional or a developer messed up, but to fix it simply set the EDC Limit to 0 instead of the maximum.

PPT and TDC can still be set normally, but make sure the EDC Limit is either at 0 or the maximum minimum your cpu uses when boosting.

Run cinebench again once you've made these changes and you'll see that your boost clocks would have returned.

Edit: setting EDC Limit to 0 just puts it at your mobo default values. so no worries


source:https://old.reddit.com/r/Amd/comments/dqwcpo/agesa_1004_pbo_bug_and_how_to_fix_it/

Yeah, saw that Reddit post too; pretty interesting. I didn't experience the issue, though. I just enabled PBO, set the Scalar to 10X and let it be. CPU Voltage reaches 1.5V, all cores boost properly. I wonder if I just got lucky or if maybe there's another setting somewhere in the BIOS that causes this bug.

 

[Sphex_]

I've built my own computers for years and worked with many gigabyte motherboards but I've never seen this kind of problem before. On my new X570 Master, everything works great until i flash the main bios to a newer version. Then it fails to load windows. However the back up bios works just fine, so it clearly not a hardware issue. I can access the main bios and change any settings but It won't load windows. It states there is no boot drive available. I thought this may have been a defective board as did gigabyte when i put in a ticket for technical support, However i got a replacement board from amazon as i thought the first was broken and now the replacement board is doing the exact same thing. It has to be a software option somewhere, but i can't for the life of me figure out what it is and have searched everywhere. Does anyone have any ideas? I've tried re-flashing the bios, clearing the CMOS, loading optimized defaults ect. but to no avail. When i go into the bios and select the boot drive i get post code is AA, which apparently stands for reserved whatever that is. If i don't go into the bios i get post code AE and a message stating "reboot or select boot proper boot device or insert boot media in selected boot device and press a key"



X570 Master, with Ryzen 3800X, 4tb Samsung 860evo Sata, and gskill 3200Mhz CL14.



Does anyone have any ideas???


Thank you.

I had a similar problem after upgrading to F5G on my Aorus X570 Elite. Flashed fine, applied my settings, booted into Windows just fine. Shut down the computer and went to bed later that night, woke up the next day and couldn't boot into Windows. No combination of CSM settings helped so I'm setting something got screwed up with the MBR or GPT Partition Table. diskpart showed no partitions at all on the C:\ Drive, so... I just reinstalled Windows 10. Definitely annoying but I don't save much on the main hard drive anyways.



Pop in a Windows 10 USB if you can and go into recovery options, launch command prompt and poke around in diskpart. See if the proper partitions are listed. Otherwise you can try the standard bootrec /fixboot and bootred /fixmbr and all that good stuff. It's worth a shot, especially if you have important data on your C Drive.

 

[Sphex_]

Assuming my drives don't disappear tomorrow, I'll try disabling CSM and test that for a day. Then I'll move onto the next setting...

Honestly, don't. Just leave things how they are for now. When I updated, CSM was enabled on my machine, but everything was set to UEFI. I had disabled CSM in the past just because I thought it was unnecessary but every time I updated the BIOS the default would re-enable it and I'd just forget to set it back after updating. Didn't make a difference so I eventually completely forgot. After updating to F5G, I disabled CSM, because I didn't need it. I saved settings, the computer booted fine, and after shutting down and turning the computer on the next day, the problem occurred, so I blamed myself. Setting CSM back to enabled, like it was before, didn't rectify the issue. But now that I'm seeing more and more complaints about this, I'm wondering if there's something else going on here. I'd be shocked if it had something to do with SATA ports or whatever. None of my other hard drives were affected either. I have a secondary SSD for games and whatnot and a 1 TB HDD for pictures, videos, other **** like that and they were fine. Although, after reinstalling Windows on my primary drive, the secondary SSD was labelled as "Foreign". Probably just a security thing. I still think it was a mistake on my part, but I have no idea how the MBR or GPT partition table could get messed up after a BIOS update after having performed a handful before hand and setting all of the BIOS values back to what they usually were.

 

[Sphex_]

Can anyone tell me why setting the "PBO Limits" value in AMD Overclocking > Precision Boost Override to anything other than Auto suddenly caps the CPU voltage to ~1.48V?

• With the PBO Limits set to Auto, like this: https://imgur.com/WiaWzz6, the CPU happily hits factory boost clocks and even exceeds it (because I enabled Auto OC), and all cores reach 1.5V maximum when boosting in certain workloads.
• With the PBO Limits set to Motherboard or Manual (raising the limits on my 3700X to the 3800X's stock value, for example), the CPU NEVER hits factory boost clocks on any core (usually maxing out at 4350 MHz), and all of the cores seem limited to around 1.48V

I suspect the factory boost clocks aren't being achieved because the core voltage is being limited in some way. Setting a positive core voltage offset does not fix the issue, in fact it seems to have no effect at all. Is anyone else experiencing this? Specs in signature.

 

[Sphex_]

Because of the infamous PBO-Bug you must set EDC to 0 - you can and should set PPT & TDC to motherboards max.(depends on your board). Set hyperscalar to 10x and use a negative offset for CPU core voltage between -0,09V to -0,10V for best results with PBO.
Also set BCLK to 100,2-101,0 MHz as this helps a lot with BOOST and shouldn’t stress all other components to much. With these settings my 3900x boosts up to 4,7GHz on my Aorus Xtreme with Waterloop that is…

Hopefully the PBO-Bug will be corrected with next AGESA...

Greetings,
Medizinmann

Thanks for the tips bro. I gave these a shot this morning and here are my findings:

• Keeping EDC at 0 seems to work. I set PPT & TDC to the stock 3800X values. They never max out anyways, and EDC, which I can't change anyway, tops out when running Cinebench, so...
• PBO + Auto OC with PPT and TDC raised works and the CPU voltage works properly. Because of this, the CPU also boosts properly.
• I'm not sure Scalar does anything, 10X hasn't made a difference, but I'll keep it there.
• Setting a negative voltage offset of even .012V adversely effected boost clocks, so I'm keeping VCORE on Auto. All cores hit 1.5V like they should.
• I can't fine tune the BCLK on my X570 Aorus Elite motherboard. It only increases in 1 MHz increments. I tried 101 MHz and SATA devices stopped being detected by the BIOS. So that's out.
• Boost Clocks while gaming are 25-50 MHz higher on average, which is nice.

Overall, it looks like all I've been able to do is raise some of the power limits. Thanks for the tip about the EDC value. I had read in the past that it was bugged and setting it to 0 would fix the issue but I misinterpreted what I was reading and thought that setting it to 0, rebooting, then setting it to the value of my choice would then apply and function properly. This is obviously not the case. With all cores reaching 4400 MHz or above, I can't really expect too much more out of it, so I'm happy with the settings I have now.

 

[Sphex_]

Hello i hope someone can help with this.

I went into my bios the other day and noticed my bios had changed to the shipped bios (f4) but i updated my bios when i built the pc 2 weeks ago to f10. so do i update the bios again or try change it. when ever i turn my pc on sometimes will take 2-3 post cycles to actually post to desktop.
3900x
aorus master x570
gskill C16 3600mhz 16gb
Win10 on 500gb 2.5 SSD

So how do i go about fixing this? is another bios more stable then f10?

Your motherboard has dual BIOS. So you initially booted to one BIOS and updated it to F10 (probably your Primary BIOS) and because of the POST cycling (memory training? Hardware issue?) it decided to boot to the secondary BIOS. Update it to F10 as well so you have F10 on both BIOSes and try to figure out why your computer can't POST properly every time you turn it on. Maybe someone with more experience with the Aorus Master board can chime in but in my experience failure to POST or power cycling is usually caused by bad memory timings or something along those lines.

 

[Sphex_]

anyone else have issues with usb 2.0 disconnecting reconnecting devices constantly and randomly?i have a auros pro wifi x570,
both my mouse and keyboard will randomly reconnect and disconnect, then ill get usb errors and vram errors, installed old motherboard and no isuues at all

Yeah I'm getting this too but it only started today. I'm still on F10C, haven't seen a reason to upgrade. I haven't changed anything with my build except for installing a USB-C front I/O port for my Fractal Design Define R6 yesterday. Perhaps this is the cause? Do you have a front USB-C connector? I'm going to try plugging my mouse and keyboard into the USB 3.0 ports on the back of the motherboard instead of the USB 2.0 Ports. Hopefully this mitigates the issue.

 

[Sphex_]

See my response right above yours

Ah, I did see that haha. I don't have HWInfo running but I do have AIDA64 running for Sensor Panel (same idea, just nicer to look at). I went into AIDA64 and disabled all Asetek and Corsair sensor related items in the "Stability" section in preferences. Unfortunately, while typing this up it happened again. It's very quick. My mouse and keyboard lose connectivity for a split second. It's very brief, but super annoying.

 

[Sphex_]

Guys. Any idea why after restarting my computer suddenly I have no boot devices. I was installing drivers and stuff and reset my pc. And now I have no boot devices. I tried putting in my USB with my install media. And according to it. Every single one of my drives are empty. And chkdsk does nothing. It's just like my install of windows just vanished. I don't know what is going on.

Ryzen 3800X
Aorus X570 Pro WiFi on Bios F11

I had a very similar thing happen to me a few months ago. Did you touch the CSM setting in the BIOS at all? I wound up having to reformat and reinstall Windows and use Disk Management to discover my other drivers. MBR or GPT got messed up somehow. All of your data is there but you can't see it.

 

[Sphex_]

PBO seems to make my voltage rather high? Is this normal or safe? My temps are good while gaming.

Totally fine.

 

[Sphex_]

Wow. Really insightful and totally relevant conversation about Gigabyte X570 AORUS Motherboards... Allow me to steer the ship away from wild, weird, Chinese promoting ramblings.

For everyone that uses Argus Monitor with their Aorus master/Xtreme, the program was updated and it seems like it can control the second ITE chip, previously the programs would read the chip but couldn't control anything. Going to rewire the fans and test Further.

How did you wind up making out with this? This is actually pretty interesting.

 

[Sphex_]

Thanks for letting me know. So is 41 when doing desktop stuff ok then?

Yes, you're fine. Mine idles around 45°C on my Aorus Elite.

 

[Sphex_]

Thanks guys, sounds like a keeper this AIO.
Amazing value for money.

Sent from my ONEPLUS A6003 using Tapatalk

How are temps with your 3600 at the speed stated in your signature?

I just installed a 280mm Liquid Freezer II on my 3700X. With PBO enabled (EDC 1 bug) and solid case airflow, I'm getting CPU temps that hover around 60°C and spike into the low 70s while gaming (Ghost Recon Wildlands, Deep Rock Galactic). Also reached about 77°C while rendering a video. I have the radiator front mounted in my case in a push/pull setup. It's incredibly quiet but I saw nearly no difference in CPU temps while gaming when compared to my Thermalright 140 True Spirit Power. Thinking about attempting a remount, but I was curious to see what kind of temps you were getting with your 3600.

I'm guessing that my chip just generates too much heat in too small of an area, when PBO is enabled in the way that I have it enabled, to be whisked away quick enough to flatten out those temperature spikes.