![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Gigabyte X570 Aorus Master won't power on - (temporary) solution
- Thread starter cold_steel
- Start date
onebitrobot
n00b
- Joined
- Oct 25, 2020
- Messages
- 5
Same here. No further cold boot issue since October 27th. I did buy new certified Display Port cables but other than that I've not changed the BIOS or any settings. Maybe it's the cables, or maybe there was some sort of Windows shutdown issue that has subsequently been patched? I've got a Asus Dark Hero on order and I'm in two minds what to do as everything has been running great for a couple of months now.
cold_steel
Gawd
- Joined
- Jun 27, 2004
- Messages
- 559
I really don’t think this ever had anything to do with DP cables.
I’m still using the exact same ones I’ve always had, and we are creeping up on 11 months since the last time I saw this issue now.
SpaceCadet2000
n00b
- Joined
- Oct 13, 2020
- Messages
- 25
November 6th was the last time for me, so also +2 months. No issues since I installed the F31j bios, and subsequently F31.
I should also note that I've taken up the habit of turning off the power supply via the switch at the back whenever I turn off my computer. Not sure if that does anything, but who knows ...
+1. No issues for more than 2 months (before I was swtitching power cable from the wall and it was 100% working workaround, now it seems not needed). Working mainly on F31 beta bioses, F31q now. Erp and "Power Loading" Enabled.
This is very odd
BTW I am not using DP cable at all, so I am pretty sure DP is (at least not always) the problem.
This is very odd
BTW I am not using DP cable at all, so I am pretty sure DP is (at least not always) the problem.
I've read complaints about the dark hero choking on memory speeds that the Master handles with ease. I'd cancel my order if I was you.
+1 again. Last time was 9th of November 2020 so 2 months for me too. I worked also on de F31 beta bioses... still on 31q... also ERP and Power Loading enabled.+1. No issues for more than 2 months (before I was swtitching power cable from the wall and it was 100% working workaround, now it seems not needed). Working mainly on F31 beta bioses, F31q now. Erp and "Power Loading" Enabled.
This is very odd
BTW I am not using DP cable at all, so I am pretty sure DP is (at least not always) the problem.
November 11th for me. I updated to the F31 bios and also have ERP and Power Loading enabled. Fingers crossed!
onebitrobot
n00b
- Joined
- Oct 25, 2020
- Messages
- 5
I'm going to throw in another random thought. The timing of my cold boot issues coincides roughly with when I got my new case and installed RGBFusion. I've seen another thread where they reckon RGBFusion writes stuff incorrectly and breaks your drives (happened to me several times!) Since I've set up my RGBs I don't run Fusion anymore. Could it be Fusion writing duff info to the mobo? Maybe not, but be aware of the (potential) broken drive partition issue if you are using it.
Sata drives being wiped
Sata drives being wiped
Last edited:
SpaceCadet2000
n00b
- Joined
- Oct 13, 2020
- Messages
- 25
Probably not. I haven't touched RGBFusion, and there are Linux users in this thread who are reporting the same issue so I think we can rule out anything on the OS/software side.
Hello,
the last time the board did not power on was on Oct 10th. It's working now over three months without issues. I never changed the cables.
On Oct 10th, i disabled Power Loading again, because it did not prevent the board from playing dead. ERP is still enabled. I also switched back to the default dual bios. The one thing which is new since the last "no power" bug is, i disabled the on-board RGB-LEDs. And i updated to F30.
Hmm, i wonder if i should risk it and enable the windows hybrid-sleep mode again... I have a 10 year old Phenom II machine runing Windows 7, last time freshly installed somewhere in 2012 when it got a SSD... and this machine boots much much faster than the "new" one with the X570 Aorus Master with Win 10 on a PCI-E 3.0 NVME SSD. That is kind of funny.
the last time the board did not power on was on Oct 10th. It's working now over three months without issues. I never changed the cables.
On Oct 10th, i disabled Power Loading again, because it did not prevent the board from playing dead. ERP is still enabled. I also switched back to the default dual bios. The one thing which is new since the last "no power" bug is, i disabled the on-board RGB-LEDs. And i updated to F30.
Hmm, i wonder if i should risk it and enable the windows hybrid-sleep mode again... I have a 10 year old Phenom II machine runing Windows 7, last time freshly installed somewhere in 2012 when it got a SSD... and this machine boots much much faster than the "new" one with the X570 Aorus Master with Win 10 on a PCI-E 3.0 NVME SSD. That is kind of funny.
The day before you asked, if I'm not mistaken, monday 11/1, and breaking with the two-month interval: hardly more than a month. We could use an electonics ace to tell us the difference between a BIOS reset and battery removal.
Fry178
n00b
- Joined
- Mar 13, 2016
- Messages
- 35
so I've been home the last couple month, and decided to narrow down the issues of the board (sudden reboots/cold boot etc), and research settings/tweak related stuff, especially since i had to build/rebuild a few rigs (3xxx + x570) in the past 2 month
5 different x570 boards from 3 different brands, similar spread for ram and psu.
please make sure that any guide/tips etc are up to date, if seen a "definite guide to.." from 2017 that was recommending the wrong things, just because stuff has changed, bios got better, win got updates etc, and the OP didnt care to update it.
couple of things posted by some, will not make any difference (most bios settings, but might actually be counter productive, like keeping a manual handbrake engaged all the time, so the car doesnt roll back when you stop at a hill once a month)..
anyone please don't take it the wrong way if i post info that differs,but ive literally spend days sitting in front of the moni testing/trying stuff, and to keep this "short" as well.
-regarding msi boards, only the cheapest x570 board isnt coming with the "best" vrm design,
but anyone buying a sub 160$ board to run a +500$ oced 12/16 core, deserves it to blow up in their face.
asus isnt the best with the ram i tested it, and i didn't like the bios (vs older amd/intel).
-the days that required certain bios settings are over, win is aware of zen2 stuff,
so leave whatever stuff is set when loading defaults, especially power loading and ccp related stuff.
if you use pbo and/or ram above 2133, you might need different settings than what comes below.
i do say use manufacturer tweaks (e.g optimized vs auto, for power related stuff),
TURN OFF ANY SPREAD SPEC related setting, use q&c, for Gb boards leave csm enabled.
regarding psu:
i stressed a psu known to be transient sensitive (650w evga G+/1) with a 3700 (pbo enabed) and an oced 2080S, and nothing triggered a (Transient load) warning (2 leds on gpu pcb) or caused a reboot, even when continuously running 550-600w (UPS load) playing fps in 4k.
yet sitting idle at desktop with office or browser open (handful of tabs, if) triggered a reboot, reproducible on a different 750w unit thats not sensitive to load changes.
-all my "sudden" reboots on no/low load where caused by ram just barely being stable, and the board fluctuating too much on dram volt.
-anyone running 1t, dont. without spending days to dial in other timings/voltage, not one rig i checked was stable with it on.
dont use memtest86 or similar,as they don't pickup "ocing" related errors, just if hw is broken.
dont consider your ram to be stable if running above jedec (2133/1.2v), no matter if its xmp or amp and qvl, unless its passing TM5 (use anta777 profile), and HCI (run multiple in parallel to fill ram) stable, and should pass 6h of high load like prime or heavy gaming to see if its unstable with higher temps.
usually turning off gear and power down, switching bgs to disabled and alt bgs to enabled, helps with ram tweaking.
- all my cold boots were bios related.
part caused by using qflash (instead of efi flash under dos) leaving data behind corrupting firmware.
after efi flashing update (which had other issues), and using efiflash to clear cmos and go back to previous one (which had been updated using only qflash) i noticed additional entries i did not have before (usb/TB related).
so using qflash to update bios (single and dual chip) was giving me cold boots, when the same update using efi flash (and /C), did not.
make sure to use /C (to clear cmos data) and /DB (for dual bios boards) with efiflash.
-unless you have a 5600 or 5800,, dont update to the latest (f30/31 for the ultra), as amd does not recommend it and can have negative impact for zen2 chips, of course nvm if you already did/F22 is causing issues etc?
and last: if it wont power on, DONT try the battery stuff right away!
unplug psu without flipping psu switch, press/hold power button (case) for 10-30s, switch the psu off,
plug cord back in and wait 5s, flip psu switch and wait 5s before trying boot again.
getting ram stable fixed my reboots, proper flashing of bios fixed the cold boots, and since then it turned on every time, so there might be a "connection", as only the versions that had cold boots (after unplugging from outlet) also produced no-power-on situations.
5 different x570 boards from 3 different brands, similar spread for ram and psu.
please make sure that any guide/tips etc are up to date, if seen a "definite guide to.." from 2017 that was recommending the wrong things, just because stuff has changed, bios got better, win got updates etc, and the OP didnt care to update it.
couple of things posted by some, will not make any difference (most bios settings, but might actually be counter productive, like keeping a manual handbrake engaged all the time, so the car doesnt roll back when you stop at a hill once a month)..
anyone please don't take it the wrong way if i post info that differs,but ive literally spend days sitting in front of the moni testing/trying stuff, and to keep this "short" as well.
-regarding msi boards, only the cheapest x570 board isnt coming with the "best" vrm design,
but anyone buying a sub 160$ board to run a +500$ oced 12/16 core, deserves it to blow up in their face.
asus isnt the best with the ram i tested it, and i didn't like the bios (vs older amd/intel).
-the days that required certain bios settings are over, win is aware of zen2 stuff,
so leave whatever stuff is set when loading defaults, especially power loading and ccp related stuff.
if you use pbo and/or ram above 2133, you might need different settings than what comes below.
i do say use manufacturer tweaks (e.g optimized vs auto, for power related stuff),
TURN OFF ANY SPREAD SPEC related setting, use q&c, for Gb boards leave csm enabled.
regarding psu:
i stressed a psu known to be transient sensitive (650w evga G+/1) with a 3700 (pbo enabed) and an oced 2080S, and nothing triggered a (Transient load) warning (2 leds on gpu pcb) or caused a reboot, even when continuously running 550-600w (UPS load) playing fps in 4k.
yet sitting idle at desktop with office or browser open (handful of tabs, if) triggered a reboot, reproducible on a different 750w unit thats not sensitive to load changes.
-all my "sudden" reboots on no/low load where caused by ram just barely being stable, and the board fluctuating too much on dram volt.
-anyone running 1t, dont. without spending days to dial in other timings/voltage, not one rig i checked was stable with it on.
dont use memtest86 or similar,as they don't pickup "ocing" related errors, just if hw is broken.
dont consider your ram to be stable if running above jedec (2133/1.2v), no matter if its xmp or amp and qvl, unless its passing TM5 (use anta777 profile), and HCI (run multiple in parallel to fill ram) stable, and should pass 6h of high load like prime or heavy gaming to see if its unstable with higher temps.
usually turning off gear and power down, switching bgs to disabled and alt bgs to enabled, helps with ram tweaking.
- all my cold boots were bios related.
part caused by using qflash (instead of efi flash under dos) leaving data behind corrupting firmware.
after efi flashing update (which had other issues), and using efiflash to clear cmos and go back to previous one (which had been updated using only qflash) i noticed additional entries i did not have before (usb/TB related).
so using qflash to update bios (single and dual chip) was giving me cold boots, when the same update using efi flash (and /C), did not.
make sure to use /C (to clear cmos data) and /DB (for dual bios boards) with efiflash.
-unless you have a 5600 or 5800,, dont update to the latest (f30/31 for the ultra), as amd does not recommend it and can have negative impact for zen2 chips, of course nvm if you already did/F22 is causing issues etc?
and last: if it wont power on, DONT try the battery stuff right away!
unplug psu without flipping psu switch, press/hold power button (case) for 10-30s, switch the psu off,
plug cord back in and wait 5s, flip psu switch and wait 5s before trying boot again.
getting ram stable fixed my reboots, proper flashing of bios fixed the cold boots, and since then it turned on every time, so there might be a "connection", as only the versions that had cold boots (after unplugging from outlet) also produced no-power-on situations.
Last edited:
SpaceCadet2000
n00b
- Joined
- Oct 13, 2020
- Messages
- 25
Not sure why you're posting all this here. It has nothing to do with the topic of this thread.
As the poster said, he or she has been home a couple of months, probably needed someone to talk to.
onebitrobot
I presume your new case has acrylic/glass side panels in contrast to your old all-metal case?
Fry178
n00b
- Joined
- Mar 13, 2016
- Messages
- 35
except that i had the problem of non power on (and issues found on other) on my ultra, and i was able to fix it, not sure how thats not relevant.
and lots of issues are not limit to the master and exist across their amd lineup, as well as some info posted by others HERE, is either incorrect/outdated, so i posted a anything that might help.
just because i will never buy a Gb board again, doesnt fix the problems some have right now, and i would still prefer to get the board to work (properly) instead of having to rip out my loop (etc) just to be able to replace the board.
(tip on the side, maybe leave it to OP/mods/admins what is or isnt OT)
and lots of issues are not limit to the master and exist across their amd lineup, as well as some info posted by others HERE, is either incorrect/outdated, so i posted a anything that might help.
just because i will never buy a Gb board again, doesnt fix the problems some have right now, and i would still prefer to get the board to work (properly) instead of having to rip out my loop (etc) just to be able to replace the board.
(tip on the side, maybe leave it to OP/mods/admins what is or isnt OT)
Last edited: