Under 100% load it drops to 1.486v. I stay below 1.55v and when gaming CPU temp is usually around 30-40c.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
*** Official Ryzen Owners Thread ***
- Thread starter Frozennova
- Start date
Under 100% load it drops to 1.486v. I stay below 1.55v and when gaming CPU temp is usually around 30-40c.
LL2 2. I could use LLC3 and get a smaller gap between the min and max voltage, but not sure it's worth it.
Max temp 59c under Prime small FFT with fans at 100%.
I replaced the stock H110i fans with these > https://www.overclockers.co.uk/cors...remium-magnetic-levitation-fan-fg-059-cs.html
I am looking to replace the stock fans on my h100i V2 actually.
I'm at 1.35v and my temps hit 55-60, I wish they would sort this temperature discrepancy out. One option in the bios throws it way off.
What are your idle temps?
Does anyone ever see weird pseudo lock ups in games or otherwise with ryzen? I was playing sniper elite 4 the other day and the system just slowed to a chug. It'd only respond every few seconds and forced me to hard reset. It would then bootloop a few times before coming back to life.
Finding it hard with the system to work out whether overclock is unstable or there are some other fundamental issues with the board/BIOS.
Finding it hard with the system to work out whether overclock is unstable or there are some other fundamental issues with the board/BIOS.
Not sure if it's your issue, but I have found certain software to not play nice with this setup - main culprits for lock-ups:
Bitdefender
Drivermax
Asus AI Suite
Known to cause DPC Latency spikes for me:
Nvidia drivers
Photoshop CC
On the latency side, this can be very much down to ram timings and procodt I have found - can test stable for hours and not be 'smooth'. lowering timings can actually help, and trying different procodt values.
Also, is this after your system has been running for a while? I still find that every now and then I can have it suddenly start feeling a bit laggy - but generally passes.
Thanks that's helpful. It has happened a couple of times and yes for no apparent reason. The other day it happened after 3 hours of planet coaster followed by 30 mins of sniper elite. It happened precisely when I turned vsync on incidentally but I've seen it a couple of other times too.
A bit more detail then -
Running 3333 at 14-14-14-34. Two hours memtest stable. 1.38v SOC at 1.125v.
Also running 3.9ghz at 1.425v, which droops all the way down to 1.35 ish under load. LLC2 didn't seem to help this at all. 1 hour real bench and 2 hours prime stable. 65c or so under said load but rarely gets over 50 in games.
However. The beta BIOS (with 1.0.0.6) doesnt like vcore over 1.25 or multiplier over x36. Going over on either causes x15.5 to be set. I get around this with 1.25v and x36 in BIOS, then have easytune ramp up the overclock and volts on bootup. Easytune feels a bit crap so wonder whether it doesn't help.
Finally I don't seem to have great airflow through the case as I am using an intake rad which doesn't need any fan speed at all to keep the CPU cool. Under load in the heatwave last week all temps were sensible apart from the VRMs for the CPU which were in the 90s under prime95. Again though nothing like that when gaming.
I am going to drop my GPU overclock completely for a while just in case. I tend to run CPUZ and/or afterburner in background on my second screen so will stop doing that too.
As well as the intermittent sluggish behaviour I also get bootloops both at 3333 and 3200. It reboots 3 times with code F6 then starts up. When it starts up all settings are fine, timings remain and there are no other issues. If I get into windows without a bootloop it will crash when easytune applies the OC. It will then bootloop, restart and be fine. I am pretty confident elements of this are due to the BIOS being in beta as once I can apply x39 I will ditch easytune...
Biggest culprit I found for boot issues is procodt.
And I also set startup dram voltage to same as regular.
Best way I found for testing procodt was when setting a value, also changing boot voltage 0.05V in either direction, as this prompts it to do a cold boot on save. Using that a few times I was able to work out that my ram/system likes 48-60 ohm.
Then, when booting into windows run tpu memtest or realbench and run for 5 to 10 min, and if you if you experience lock-ups for lengthy periods. 53/60 seem to work best for me.
Haven't had any boot issues, and thats running my 3000Mhz 64GB at 3066.
And I also set startup dram voltage to same as regular.
Best way I found for testing procodt was when setting a value, also changing boot voltage 0.05V in either direction, as this prompts it to do a cold boot on save. Using that a few times I was able to work out that my ram/system likes 48-60 ohm.
Then, when booting into windows run tpu memtest or realbench and run for 5 to 10 min, and if you if you experience lock-ups for lengthy periods. 53/60 seem to work best for me.
Haven't had any boot issues, and thats running my 3000Mhz 64GB at 3066.
Is that using 4 sticks of ram eyetrip ?
Yes, 4x16gb.
Also have to run soc voltage around 1.2V for this - even at 2400Mhz it could result in display cut-out under load.
Running 4 8gb sticks of Teamgroup Pro Dark 3200 here at the rated speed and timings of 16-16-16-36. It did take some time though to find the sweet spot for procodt, in my case 53.3ohm. Once i found it though it stopped all the cold boot issues straight away, not had one issue booting up since. Soc is still on auto and showing as 1.15v. I'm guessing you need 1.2v because of 64gb double sided ?
I get displaycut out during memtest - is that SOC related? Just thought it was a quirk lol.
Rapidly discovering this system isn't as stable as I thought it was
I have been clocking since the 1990's and to be honest i have never found ANY Windows software based clocking app very reliable. I'd suggest doing any clocking in the bios and then stress test in windows. That way you can rule out software affecting your clock.
I disabled Sensi MI Skew and my CPU temps went up 10C or so. Not sure which is the right setting tbh.
Looks like Asus added a temp offset with bios 1401.
Idle temps were 20-30c, now up 10c.
Also, do you have any gpu overclocking software?I get displaycut out during memtest - is that SOC related? Just thought it was a quirk lol.
Rapidly discovering this system isn't as stable as I thought it was
If so, uninstall/disable if soc voltage don't help.
I'd have thought 10 higher would make more sense at your voltage.
That would seem more like it. I'm at 30c idle with 1.35v