Aorus X570 Master

[Calmmo]

x590 could be the zen3 out of the box ready moniker for x570

 

[jesdals]

Setting PBO to auto and disabling Core Watchdog seems do give better boost - will try with fixed cpu vcore next week - but stable so far and optimized for mem as F12g

Average cpu speed seems more or less the same, 73,3c chipset temps is a challenge cant get them under 70c in intens load. Wonder if the B550 master has better temps

 

[TheLostSwede]

x590 could be the zen3 out of the box ready moniker for x570

No, according to what I've been told, is the naming AMD uses and Gigabyte hasn't change it. So far, there's no X590 chipset.

Setting PBO to auto and disabling Core Watchdog seems do give better boost - will try with fixed cpu vcore next week - but stable so far and optimized for mem as F12g
View attachment 161667
Average cpu speed seems more or less the same, 73,3c chipset temps is a challenge cant get them under 70c in intens load. Wonder if the B550 master has better temps

Surprised you're not seeing even higher boosts, as I get up to 4,575MHz peak boosts on a couple of cores. I guess your air cooler might be what limits you a bit.
That said, it's not like an extra 50MHz peak boost would make much difference in 99.999999% of cases.

 

[jesdals]

I havent been giving specific vcore voltage on the replacement board yet - before I used to set it between 1.4 and 1.4125v. 4.525,3 MHz seems to be the limit with current settings and only on 3 cores 3 more at 4.500MHz and 2 cores at 4.475MHz

 

[TheLostSwede]

I havent been giving specific vcore voltage on the replacement board yet - before I used to set it between 1.4 and 1.4125v. 4.525,3 MHz seems to be the limit with current settings and only on 3 cores 3 more at 4.500MHz and 2 cores at 4.475MHz

I haven't touched the Voltages on my board.
Sometimes you need to leave hwinfo running in the background to hit the peak speeds, it seems to happen more frequently when gaming, rather than when you try to coax the CPU into hitting peak clocks...

 

[Divide Overflow]

Why do the AGESA versions seem to be going down?
F12g was 1.0.0.4
F20b is 1.0.0.2

 

[heky]

Why do the AGESA versions seem to be going down?
F12g was 1.0.0.4
F20b is 1.0.0.2

They are not, F12g was Agesa V1 1.0.0.4, F20b is Agesa V2 1.0.0.2

 

[jesdals]

Did som testing with cpu vcore set to 1.2v and 1.3v, at 1.2v it was not stable enabling Hardware Info made it reboot. But at 1.3 its stable

It seems it caps my boost to below 4.5GHz but average is more or less the same, chipset temps still high after 5 ours of Division 2. But CPU temp has not been above 62c thats 8c lower than with auto setting of vcore. But the vcore max has stil been at 1.5v - that 0.2v above the bios setting and the same as on auto. Will do some more testing at these settings before raising it to 1.4v and 1.4125v. Might test the all core pstates OC at 4525MHz again as well.

 

[Assimilator]

They are not, F12g was Agesa V1 1.0.0.4, F20b is Agesa V2 1.0.0.2

It's AMD's latest and greatest way to unnecessarily confuse consumers.

 

[jesdals]

Gamers Nexus did some 3600XT tuning on the Master and show several settings on memory etc.

Setting VDDG stil dosent bring Infity fabric above 1866 for me, so I live those at auto. Still on F20b bios.

But setting the load line calibration to high seems ok, stil at 1.3v vcore and are going to test that for a couple of days still. But have managed to get memory more tight

For some reason Trcd is the most tricky, but with 4 sticks these settings are stable, have tried Trfc at 240T and below but bring stability issues. Geardown mode is enabled and it sets the Command Rate at 1. CPU vcore still goes high at 1.3v and my chipset temps also goes above 75c. Will try to raise vcore though to see if it grants better boost.

 

[jesdals]

Done some more testing with the F20 bios and its seems stable - BUT if your using SATA drives do read this thread first https://www.techpowerup.com/forums/threads/crazy-gigabyte-x570-master-issue.260727/#post-4311777

Well been testing with cpu vcore at 1.3v and mem at 1.5v and found the settings stable been playing arround with the load line setting it to normal - on my first board I could not tuch that setting without it being unstable, but even normal grants the boost and settings above. Yesterday I saw the new Buildzoid rambling about lovering DDR VTT

and decided to try it out on my settings sat them to 0.5v and it ajust it as shown below to 0.660v ish. Edit setting DDR VTT to 0.5 did not enable more tight mem settings tinkering around but no succes with CL13 or below.

Today I have been testing Pstates all core settings at 4475MHz its only stable if I raise cpu vcore to 1.393 volt - but If you compare with the auto settings above, its strange to se that the individual cpu vcore never goes above 1.393v where as on the auto setting it does go all the way to 1.5v

The 4475Mhz is below the 4525MHz I had it running at on the old board, but doing all these testing I have found that all 8 core can do 4475MHz but will try to raise the all core later. I will also try to se if lowering DDR VTT is going to enable tighter timings. Buildzoid used it to raise the speed, but I think thats pointless on Ryzen.

Did some more testing - it seems that the chipset gets a bit hotter - but look at that vcore for the individual cpu cores

it seems strange that i does not go above 1.1v

 

[heky]

it seems strange that i does not go above 1.1v

But it does, look at the Cpu Core Voltage or VR VOUT reading...you are running between 1.32v and 1.42v. Ignore the VID readings.

 

[jesdals]

Been testing both Pstates all core OC and the auto setting with PBO on auto - the result a all core OC at 4475MHz does not give any difference in Division 2. Result below is the average result out of 5 runs and the difference is marginal in favor for leaving CPU at auto with PBO on auto in bios

After a bios reinstall I could not remember if I used to have CPU virtualization on - apperantly no difference in above and below result

The all core results above does not differ much from below auto setting of the CPU with 1.3v vcore on cpu

A genral difference at only 20-30 points but stil with higest hig at 3267 point and lowest high at 3240 point. I see no advantage of all core OC. Results might differ with higher OC, but then you need to go above 1.4v vcore - I have had this CPU above 4600MHz with 1.5v+ vcore but bot 24/7 stable. Result will differ with other games but this is my average use.

 

[TnF]

Hello. My X570 Aorus master rev1.0 (#~4200) is dead.

PC was working perfectly fine and was not even on a load, suddenly it froze and hard rebooted...went into desktop after the reboot, froze and hard rebooted a 2nd time with dram led error end code 00.

Troubleshooted all the parts cpu/ram/psu/gpu and it's the mobo, bios doesn't make a difference either. I did a full disassembly and clean with iso-alcohol and still same thing.

Thinking of RMA if they at least send me a new last rev board, or i will return it to Amazon and buy something else.

 

[jesdals]

If both bios dosent make a difference and you have testet the rest as you mention its RMA time. When my first board died it did not even show post. Btw what CPU do you have?

 

[TnF]

Yes mine doesn't post either, it doesn't go past the 4 debug leds so qcode is 00. 3950X is what i use with noctua nh-d15. After i read the horrible RMA stories by Gigabyte i contacted Amazon to see if i can return it to them instead.

 

[TheLostSwede]

Well, my board died last night...
Shut down the system after having used it all day, came back 20 minutes later, as I was going to look something up and it wouldn't power on.
I don't even get any digits on the debug display.
Onboard power button lights up, that's it. Press it and it goes dark.
I guess I'm fortunate enough to know some people at Gigabyte, but still frustrating.

 

[Chomiq]

Looks like there's some serious design flaw somewhere if those boards are just dying randomly. Glad to have my Elite working fine as it is.

 

[mstenholm]

Well, my board died last night...
Shut down the system after having used it all day, came back 20 minutes later, as I was going to look something up and it wouldn't power on.
I don't even get any digits on the debug display.
Onboard power button lights up, that's it. Press it and it goes dark.
I guess I'm fortunate enough to know some people at Gigabyte, but still frustrating.

Sad for sure. It would be interesting to know if the deaths are due to a common reason.

 

[Caring1]

Well, my board died last night...
Shut down the system after having used it all day, came back 20 minutes later, as I was going to look something up and it wouldn't power on.
I don't even get any digits on the debug display.
Onboard power button lights up, that's it. Press it and it goes dark.
I guess I'm fortunate enough to know some people at Gigabyte, but still frustrating.

PSU checks out alright?
It just sounds similar to what happens when you turn power off to a board at the PSU and the Power light remains lit until you press it.

 

[jesdals]

Well, my board died last night...
Shut down the system after having used it all day, came back 20 minutes later, as I was going to look something up and it wouldn't power on.
I don't even get any digits on the debug display.
Onboard power button lights up, that's it. Press it and it goes dark.
I guess I'm fortunate enough to know some people at Gigabyte, but still frustrating.

Sounds more or less like my experience with the first board - if my current dies I would consider the B550 version instead - hope you have a speedy rma

 

[TheLostSwede]

PSU checks out alright?
It just sounds similar to what happens when you turn power off to a board at the PSU and the Power light remains lit until you press it.

I presume so, I don't have anything to test it with, but it's less than a year old. There's power to USB devices, as LEDs etc. light up and I have a pair of USB speakers that detects that they're connected (as they auto power on).

Edit: Shorted the pins on the PSU and it powers on and powers up the AIO cooler, so it should be fine.

Sounds more or less like my experience with the first board - if my current dies I would consider the B550 version instead - hope you have a speedy rma

Well, I live 15 minutes from Gigabyte's HQ and as I said, I have a few friends that works there, so I'm not even worried about that part.

Sad for sure. It would be interesting to know if the deaths are due to a common reason.

I'll let you know if my mates there tell me what they find.

 

[jesdals]

Did a all core AIDA test versus PBO and auto settings, previus at 4500MHz all core

View attachment 138172
Well a small bit better in F10 bios, had to reflash because my try for 1900Mhz made me brick one bios - fortunately - F10 allows for dual flash of bioses - but did feel a bit dangerous

Pstates at 4475MHz

And with PBO and auto cpu settings

Well do find it at bit disapointing that my trimmed CL settings dosent show better performance than previous CL 15

 

[TheLostSwede]

Did a all core AIDA test versus PBO and auto settings, previus at 4500MHz all core




Pstates at 4475MHz
View attachment 164110

And with PBO and auto cpu settings
View attachment 164113
Well do find it at bit disapointing that my trimmed CL settings dosent show better performance than previous CL 15

So jealous of your CPU and RAM

3800/1900MHz not an option or you get higher latencies that way?

 

[jesdals]

well mine never did 1900 IF its only 1867 for me. Did hope the new board would grant it, but not possible. BTW could you perhaps ask your friends at Gigabyte about the difference between setting multiplier and using the Pstates setting, it seems that all people use the multiplier setting for all core OC - but the Pstates setting is so easy - so whats the difference? If any

New F21 bios - doing testing - and no not able to run IF 1900MHz on this one either

Did a test with the same PBO settings and I am running with the 1usmus Power Plan settings in bios as well

Seems stable at the same settings as F20, do not have any SATA so please be aware that there was people with SATA problems on F20 bios