MSI's X570 Tomahawk May Become Go-To OC Board??

GregI · 15 Feb 2021 at 13:20

Woodsta888 said:
IS the 8pack stuff that temperamental that it wont work at XMP?

They shouldn't be. When you said you ran memtests did you run say testmem5 for a couple hours atleast? If you only ran memtest86 (not recommended btw) then more tests are needed to see if you just got a bad kit. If you need it: https://www.overclock.net/threads/memory-testing-with-testmem5-tm5-with-custom-configs.1751608/

Yaayuh! · 15 Feb 2021 at 21:03

Yaayuh! said:
I'm getting on well with this board. Straight off the bat it's boosting higher than my CH6 on lower core usage, all-core is roughly the same, though trying to figure out best PBO settings.
RAM overclocking on the other hand is a bazillion times better on this board.

Being cheap though the USB layout on IO and lack of clear CMOS button is disappointing, but i'm used to it now.

Spoke too soon, RAM overclock is about the same, just less effort required sorting out drive strengths and the like.

Had this RAM (4x8GB GSkill 4000MHz CL18) for quite sometime now, it’s not great but it’ll do until I jump to DDR5. Currently sitting at 3600 CL16.

Fab7 · 16 Feb 2021 at 12:21

GregI said:
If you only ran memtest86 (not recommended btw)

Why ? I use Memtest86 on bootable USB key , full test it takes more than 6 hours for 2 x 16 Go , it's not the best way ???

GregI · 16 Feb 2021 at 15:52

Fab7 said:
Why ? I use Memtest86 on bootable USB key , full test it takes more than 6 hours for 2 x 16 Go , it's not the best way ???

memtest86 is like asking your RAM to do 2+2. Even if it's unstable it will most likely do that. I've seen so many people pass a test like this and then fail actual stability programs, or even while gaming will get crashes. In my experience it also doesn't seem to care if you have an unstable fclk.

I would recommend testmem5 and if you want to avoid Windows corruption you can run like 2 passes of memtest86 via USB to make sure you haven't completely messed something up with BIOS settings. Then you can start the actual stability testing. Ideally you would want hci memtest deluxe (bootable usb) to replace memtest86 but I understand most users don't want to pay for memory testing software.

Fab7 · 16 Feb 2021 at 23:39

Thanks for the informations i will try it

EDIT : i am fine ^^

========= TestMem5 Log File =========
Customize: Extreme1 @anta777
Start testing at 3:21, 2.8Gb x8
Testing completed in 2:33.26,
no errors.

philo-sofa · 17 Feb 2021 at 15:46

Yaayuh! said:
Spoke too soon, RAM overclock is about the same, just less effort required sorting out drive strengths and the like.

Had this RAM (4x8GB GSkill 4000MHz CL18) for quite sometime now, it’s not great but it’ll do until I jump to DDR5. Currently sitting at 3600 CL16.

I'm not a huge fan of the lack of clear CMOS button or lack of auto booting into safe mode. However I did buy a £2 standalone power switch/cable and hook it into the clear CMOS pins when I'm overclocking; saves me having to short pins manually.

Jimbo · 19 Feb 2021 at 11:46

Sorry if ive missed this but I have Teamgroup DDR4 Pro Dark 32gig 3600mhz CL 16 16 16 38 1.35v ram. My problem is i just go into bios on this board and enable XMP 1 and it boots fine @ 3600mhz. But if i enable resizable bar and above 4g decoding the board just does not even display anything at all. I have to re set bios with a jumper which is a pain. Any ideas what im doing wrong? I have a 5600X and 6800XT also but i suspect its something to do with ram.

Forceflow · 24 Feb 2021 at 22:30

New drivers posted on the support page for Bluetooth and Wifi somehow are older versions than the ones before:

BlueTooth Driver
22.10.0.2
2021-02-23
Intel WIFI Driver
22.10.0.7
2021-02-23

For both those drivers, the previous zip files contained 22.20.x versions.

4K8KW10 · 25 Feb 2021 at 06:56

Forceflow said:
New drivers posted on the support page for Bluetooth and Wifi somehow are older versions than the ones before:

BlueTooth Driver
22.10.0.2
2021-02-23
Intel WIFI Driver
22.10.0.7
2021-02-23

For both those drivers, the previous zip files contained 22.20.x versions.

What's the point of this? Are they saying that the older drivers are better, more stable and should be installed?

Forceflow · 25 Feb 2021 at 19:00

4K8KW10 said:
What's the point of this? Are they saying that the older drivers are better, more stable and should be installed?

Dunno. Sent in a support ticket, and today they reverted to the 22.20.x versions (which were online in january already).

hyperseven · 27 Feb 2021 at 22:44

now showing drivers from 25th

GregI · 5 Mar 2021 at 08:09

1.2.0.1 available.

https://www.msi.com/Motherboard/support/MAG-X570-TOMAHAWK-WIFI

As posted in other thread, L3 cache fix and some other stuff.

JRS · 5 Mar 2021 at 11:29

GregI said:
1.2.0.1 available.

https://www.msi.com/Motherboard/support/MAG-X570-TOMAHAWK-WIFI

As posted in other thread, L3 cache fix and some other stuff.

Nice, will try it out.

Inb4 4K8KW10's 'oh noes, illegal beta BIOS!!!111oneonetwothree'

hyperseven · 5 Mar 2021 at 11:42

JRS said:
Nice, will try it out.

Inb4 4K8KW10's 'oh noes, illegal beta BIOS!!!111oneonetwothree'

BETA, nooooooooooooooooooooooooooooooooooooooooooooooo

JRS · 5 Mar 2021 at 12:37

Okay, quick and dirty test. I ran a SiSoft Sandra 'Overall Memory Score' benchmark before and after upgrade (and after putting all my overclocking and memory tweaks back in, c'mon motherboard manufacturers there has to be a better way!). Aggregated score under Cache and Memory Latency went from 60ns to 54.2ns.

***edit***

CB R20 up by a hair - 4697 to 4734. Calling that 'margin of error'. R23 multi basically unchanged at 12010.

***edit 2***

Interestingly, it will actually boot at 4000MHz CL16 RAM, 2000MHz FCLK/UCLK now. It's not stable at that (GSOD and restart once in Windows), but it wouldn't even make it to the Win10 login screen before.

Alex h · 5 Mar 2021 at 15:54

JRS said:
Nice, will try it out.

Inb4 4K8KW10's 'oh noes, illegal beta BIOS!!!111oneonetwothree'

Lol

SpudMaster · 5 Mar 2021 at 22:46

JRS said:
Okay, quick and dirty test. I ran a SiSoft Sandra 'Overall Memory Score' benchmark before and after upgrade (and after putting all my overclocking and memory tweaks back in, c'mon motherboard manufacturers there has to be a better way!). Aggregated score under Cache and Memory Latency went from 60ns to 54.2ns.

***edit***

CB R20 up by a hair - 4697 to 4734. Calling that 'margin of error'. R23 multi basically unchanged at 12010.

***edit 2***

Interestingly, it will actually boot at 4000MHz CL16 RAM, 2000MHz FCLK/UCLK now. It's not stable at that (GSOD and restart once in Windows), but it wouldn't even make it to the Win10 login screen before.

Sounds promising. What is the pbo cpu boosting like this bios. Is it same as current stable bios or gone better again like previous betas? My cpu would boost 100 - 200mhz higher on previous betas!

JRS · 6 Mar 2021 at 10:38

SpudMaster said:
Sounds promising. What is the pbo cpu boosting like this bios. Is it same as current stable bios or gone better again like previous betas? My cpu would boost 100 - 200mhz higher on previous betas!

Identical boost to v150 stable.

JRS · 9 Mar 2021 at 15:19

This is what AIDA64 gives with the current beta BIOS on my setup:

Anyone running a similar config with the previous BIOS release who can give figures for comparison? If not, I'll re-flash v150 stable sometime to test.

bris · 9 Mar 2021 at 19:52

This board is driving me nuts. Saw a new bios for improved ryzen performance so who wouldn't want that right?
Wrong followed the M flash to the letter, went into bios, got the usb stick, loaded the bios and hit mflash. Just went to black screen and now cant get it to post.
Took out battery, cleared cmos, nothing, reflashed with the latest stable bios, which was the one I was already running and still nothing. Tried 1 stick of ram in all the slots, took out all other drives but still just
Black screen with the no bootable drive found light on.

Any one any ideas? Just cant get it to post a bios screen.