Soldato
Doing that now - 25 minutes unplugged without the battery enough do you think? I've shorted the JBAT1 header too.
MSI's X570 Tomahawk May Become Go-To OC Board??
- Thread starter DebbieY
- Start date
Should be plenty. I usually do 10 mins, but have left it longer before.
Soldato
So, I'm all sorted now... I successfully used the bios button on the i/o shield, round the back of my rig. I used a USB 3.0 Kingston Datatraveller - which I've used countless times before yesterday without issue, although updating in the usual way through the bios.
One thing I don't think any of the MSI videos/guides make very clear, well, at least not to me, was that you need to have the machine powered off when you start the flashing process. I was connecting the stick before powering on and then pressing the bios button with the system running! Lesson learnt
Thanks all! The MSI util that greats you in Windows after updating was quite useful/insightful.
One thing I don't think any of the MSI videos/guides make very clear, well, at least not to me, was that you need to have the machine powered off when you start the flashing process. I was connecting the stick before powering on and then pressing the bios button with the system running! Lesson learnt
Thanks all! The MSI util that greats you in Windows after updating was quite useful/insightful.
So other than @slipd ‘s trials and tribulations, any one else have any success or disaster stories from this bios update? I’d be keen to know if anyone can absolutely confirm that the EDC ‘bug’ is gone and we can see the clocks being achieved back to the sort of levels that ppl were getting at the start of this thread!
Soldato
I've not installed the latest BIOS yet but there's reports on Reddit saying that the latest AGESA 1.2.0.A fixes the EDC bug
Soldato
EDC bug?
Firmware from about the middle of last year created an issue whereby if you increase EDC above 140 (I think, I’m doing this from memory, and that number is for 5900x, I have no idea about the other cpus) then voltage per core on an all core load was throttled. So you could be well under the heat limit and inside the power budget for the other two values but voltage per core wouldn’t go higher then about 1.3. So when at boost you were throttled big time. Apparently this is now gone so you can up EDC to whatever you want and you get throttled now only by whatever limit your chip hits.
my Cinebench scores have already gone up nicely How new is the BIOS? I've not got to the bottom of my PC occasionally stuttering
Soldato
Not done any proper testing yet but my PC seems to now be unstable with the latest BIOS. Noticed games would freeze for a little bit in the menus but then seem to recover and work normally. Ran Prime95 and one of the threads failed within about 5 minutes. Running exact same settings as with previous BIOS Gonna test properly tonight to see if it's CPU or RAM. Could just be due to the hotter weather though I guess but CPU temps did look pretty similar to previously, maybe RAM is getting hotter now....
I backed off my ram Oc (previous set to the safe timings for 3600 with Ryzen dram calculator) back to normal XMP and all stability issues went away. I then set it back to a very mild oc (1t, 3600, CL 14 and 15 and a few others) and it seems fine again.
One thing I also noticed is that I can push much lower on my top 4 cores with curve optimiser now, which is a nice bonus
One thing I also noticed is that I can push much lower on my top 4 cores with curve optimiser now, which is a nice bonus
Last edited:
Soldato
After a quick test last night it appears the latest bios has made my RAM unable. Gone back to basic XMP timings and its stable again. I didn't check what SOC voltage HWMonitor was reporting under load but in BIOS it was reading the same as previous BIOS. I'll do more testing over the weekend but in all honesty I've not noticed any difference going back to XMP. Boot time is 0.5 second longer and 3DMark Timespy and Firestrike scores dropped a few hundred points when when you look at fps scores for each test there was barely 1fps difference.
Last edited:
Did you check the other voltages? When I put my 5800x3d in my unify the memory voltages all went down 0.05 volts I believe. I also had to adjust for the big where the motherboard give too little power to the chip...the one you test for with hwmonitor - I can't remember the name of it
Read this
www.overclock.net
Before I did this I couldn't run -30 on my 5800x3d - after I could
Read this
MSI MEG X570 Unify Overclocking & Discussion Thread
www.overclock.net
Before I did this I couldn't run -30 on my 5800x3d - after I could
Last edited:
Soldato
Any one else getting slow read and write speeds from the second m.2 slot? Not sure how long this has been the case for me because when I got drive I test with CrystalDiskMark and speeds where fine. Don't remember exactly but ~7000MB/s read and 5500MB/s write. Tested lastnight and I'm now only getting ~3200MB/s both read and write. Running latest BIOS and firmware for the drive. WD Dashboard reports the drive is using PCI-E gen4 and the drive is only 50% full.
Wondering if any of the later BIOS's broke something or if there's a BIOS setting I need to check.
I know WD SN850 NVME drives have had issues with X570 boards when running in the m.2 slot off the chipset but my understanding is that this was fixed by a firmware update for the drives and it was working at full speed for me previously
Wondering if any of the later BIOS's broke something or if there's a BIOS setting I need to check.
I know WD SN850 NVME drives have had issues with X570 boards when running in the m.2 slot off the chipset but my understanding is that this was fixed by a firmware update for the drives and it was working at full speed for me previously