Help with X570 Master bios settings?

[Mikeescott]

Would anyone here be able to help me with very stable BIOS settings on a gigabyte x570 motherboard with AMD 3950x processor? I do not need it over clocked any because I built it good enough for everything I need. All of my previous computers have been Intel extreme processor builds, and I get about three to five years use out of each build.

My build consist of a GIGABYTE X570 AORUS MASTER motherboard, AMD RYZEN 9 3950X CPU, 2 X G.SKILL 64GB 2X32 D4 3200 TRZ RGB RAM kits (128 BG), AORUS GeForce RTX 2080 Ti XTREME Graphics Card, Seagate Firecuda 520 2TB SSD PCIe Gen4 NVMe (OS drive), 2 x Intel 660p 2TB SSD PCIe Gen3 NVMe (storage drives), and EVGA Supernova 1000 T2 80+ Titanium, 1000W PSU.

This was my first AMD build, and I completed it about six months ago. Earlier this week, Windows 10 File Explorer could no longer see the 2 Intel 660 P drives that were in RAID 0 as a 4 TB D: drive that I had 3.8 TB if DATA on from school and other data/downloads. I was able to recover the data and reformat the drives in non RAID so I could use them again.

For the past two weeks I have been getting blue screen of death every single day after about it 2 or 3 hours of schoolwork and usage.

[GigaBusterEXE]

what cooler do you have?
was my first thought

maybe one of the drives are going bad?

[Mikeescott]

On the motherboard, I’m using EK water cooling on the mono-block, GPU, and chipset. I have large heat sinks on all three of the NVMe drives. Everything stays around 32 to 38 degrees Celsius. The drives are all just fine. After I reformatted, everything works perfectly. I’m looking for completely stable settings for the BIOS so this doesn’t keep resetting constantly. These settings need to include timings for the memory as well.

[Mikeescott]

[GigaBusterEXE]

I like to manually put in 1.4v on DDR4 even though most are 1.35v or lower, its 24/7 safe, no degradation
at work I’ll post more later

[Mikeescott]

Not sure if it will help, but here is my current hardware info sensor info.

[Mikeescott]

[Mikeescott]

[GigaBusterEXE]

noticed a few things, your Power reporting deviation is really off, and the power limits seem pretty saturated for idle

that’s my amateur opinion, there are peope that probably know better than I would about these topics

my next guess was chipset temps but that’s fine as well

it could be just one of your sticks of ram is bad, a thorough memtest would be worth a try

[Mikeescott]

This should explain the deviation:

w dot hwinfo dot com/forum/threads/explaining-the-amd-ryzen-power-reporting-deviation-metric-in-hwinfo.6456/

Memory is only 6 months old and never tests bad. I will run a deep scan after I’m done with homework.

[Mikeescott]

That took quite a while, but here are the results from the memtest86:

[Mikeescott]

[MEM ERROR - Data] Test: 7, CPU: 20, Address: 590EC5F8, Expected: 7FFFFFFF, Actual: 7FFDFFFF

MtSupportRunAllTests - Test execution time: 3299.604 (Test 7 cumulative error count: 5)

[MEM ERROR - Data] Test: 5, CPU: 20, Address: 590EC5F8, Expected: 20BB6E41, Actual: 20B96E41

MtSupportRunAllTests - Test execution time: 1105.897 (Test 5 cumulative error count: 1)

[MEM ERROR - Data] Test: 7, CPU: 20, Address: 590EC5F8, Expected: FFFFF7FF, Actual: FFFDF7FF

MtSupportRunAllTests - Test execution time: 3299.906 (Test 7 cumulative error count: 7)

[GigaBusterEXE]

@wendell is this a normal error rate for RAM this dense
I’ve never actually encountered errors before in a MemTest

[wendell]

No it shouldn’t error, but 128g is technically an overclock to have memory running this fast.

Disable geardown mode? And maybe back the speed off to 2933?

[Gnuuser]

also make sure the memory is seated tightly !
even the best mem-sticks may have a sloppy lock notch.
and as wendell said back the speed off a bit
38 c (around 100 f) is a bit warmer than i would like in a system
memory is susceptible to heat but often is not cooled using a liquid cooling system.
installing a cooling fan to move the air around the mem sticks could help a lot.
there are quiet fans out there that would work for it.
generally i try to keep systems below 30 c if possible.
some newer systems have temp control settings in bios that will boost the speed of cooling fans to compensate for excessive heat buildup.

[Mikeescott]

Do I need to turn off XMP? I will go in and disable geardown mode, as well as lower the speed to help. I have a fresh intake fan just below the RAM and an exhaust fan just above the board, so fresh air is flowing in from the bottom of the case and up across the MB and out the top of the case. I know these AMD CPU’s run hotter that Intel, but the liquid cooling should take care of that. There is a 520mm radiator for the CPU alone.

[Mikeescott]

OK, I turned off XMP and gear down and got zero errors on the memtest.