Stable overclock now can’t be touched

JamGeo

JamGeo

JamGeo

Associate
Joined
18 Jan 2019
Posts
82
Location
Stoke-on-Trent
After some information on an overclock on my machine. I have a 9700k. Z390 Aorus pro motherboard and some Corsair 2x8gb ram which is on the motherboard list and is up to 3000mhz.

it’s been fine running with a little bump on the CPU to 4.9ghz and the RAM using similar XMP settings to get it to 3000mhz. So the oc was nothing crazy, but a nice little boost

I have recently added another NVME and removed my sata ssd. So now I don’t have any sata drives but have 2 NVME drives. Ever since I have added this extra I’m unable to even add the overclock to the RAM without getting boot failure detected on load. I’ve been able to add the cpu overclock to 4.8 but have had some BSOD’s with WHEA_UNCONTROLLABLE_ERROR.

My question is. Would the NVME be causing the pc to no longer add an overclock? Or would it be something completely different?
I have a 750w rmx Corsair psu, so I feel this can be ruled out, I’ve also not had the system power down with a power failure. Any help appreciated.
 
pc-guy said:
what volts you using on CPU and RAM - high volts could have degraded the chip
did you test your system after the initial OC - if not there could be some inherent instability in there which has become more prominant as you have been running unstable system for a while and the cascade effect is coming afro.

there could also be some issues with power delivery system - PSU and mobo.

but start with CPU, RAM diagnostics. if none of those are at fault then go down to PSU and mobo by swapping them out.
Click to expand...

I’ve had my CPU on 1.275v to get it to 4.9 and that was running at that for a while. I did take off XMP and add my own pc to the ram which originally had instability so I did knock it closer to what XMP was doing.

mad an update I’ve been able to get my CPU back to 4.9ghz but the moment anything is added or changed to the ram I get boot failure detected which is leading me to this now being a RAM issue?
I have a cmos battery coming today to rule it being an issue with a dying battery but I don’t think that will have any effect on it
 
pc-guy said:
i think your ram has issues by the sound of it. what is the rated speed of that ram? have you checked the ram slots ie putting ram in the different set of slots and also tried 1 stick at a time.

when testing ram wiht memtest, do 1 stick at a time as well and if error found, go to the next recommended ram slot and test again until you done all 4 slots. if still error found then you got a bad stick, if no error found then you know your motherboard dim slots are ok and that stick of ram is ok. then you can test the next stick in the recommended slot without needing to swap it into other slots.

if both stick works and when they both slot in the motherboard and cannot run at XMP settings then the CPU might be bad or you haven't yanked enough volts into it. so try upper the volts especially on the memory controler (i dunno if intel has such volt).
Click to expand...

I think you may be correct with it being the RAM. The ram is Corsair 3000mhz

I cleared everything in the BIOS and started fresh. First ran memtest and that was fine but that was on optimised defaults (not sure if it applied the 3000mhz here as it wasn’t set in bios).

Next I managed to get a 4.9 and a 5.0overclock on my cpu. That boot fine and didn’t give me any issues.

After a lot of trial and error with adding XMP or manually boosting to 3000 for the ram.
I moved to the route of testing individual sticks without my cpu oc and these would still not boot 3000mhz for the ram.
I decided to to a manual pc to 2666 using the same settings that XMP uses for the oc (15-17-17. 35. 52 1.35v) on one stick at a time. Low and behold both sticks by themselves in slot 2 boost to 2666 with my cpu at 4.9. I got everything back in and now both sticks are at 2666 and CPU at 4.9ghz. I played wow, overwatch and some watchdogs legion without any issues last night, which has got me on the route of it being probably something I set on my ram before I got the NVME drive which has fried it a bit and now can’t hit 3000. I might look for some 3200mhz cheap in some sales but I’m glad I got some oc onto it and narrowed it down.

does that sound like it would be the ram? The very odd boot has got to 3000 but when I restart the pc gets in a boot loop then goes back to the last good bios setting
 
pc-guy said:
right after you cleared CMOS and booted into bios and loaded "Default" or "Optimised Default" the CPU should be on auto (no manual OC) and the ram should be on its base rating ie 2666 or 2333 MHZ whatever it is.

Boot and run memtest. if error ram is broken. but test 1 stick at a time like how i described previously to rule out any issues with the DIMM slots

if memtest is ok on all default. go in bios only load up the XMP ram setting but you need to change the Vram to 1.35V as default will be 1.2V or 1.25V. you probably need to change some other voltages on the CPU side (i am not too sure what they are on the intel platoform but the same voltages you need to get OC) but DO NOT oc you CPU at this stage. just let the CPU do its own thing.

boot and test ram together and let the process finish. if you are unable to boot then yes RAM is the issue. RMA. if you find errors in Memtest then RMA.

etc
Click to expand...


Seems to be some strange behaviour on this. I’ve not gone down the memtest route on each dim yet, but I will do. I had a bit of time this morning so I went back to default settings then got XMP in slot 2 on both sticks, this seemed to get it to work. I then moved to slot 4 which didn’t work and all other slots wouldn’t work for any XMP.

So I reset, then slot 2 wouldn’t work for XMP which had me scratching my head at this point.
So now I’ve reset everything, manually added 3000mhz and 1.35v to dram to 1 stick in slot 2. Then added a 4.8ghz oc to the cpu. Then added the second stick to slot 4. This has booted at 3000 with manual overclock to the ram. So I’ve upped the cpu to 4.9 and now I’m sat with both the ram at 3000 manually and my cpu at 4.9 which is where I want it.

mice switched off the psu at the wall and done a few resets and shutdowns and it’s continued to boot just fine.

I have no clue if there was an issue before or if at some point this oc will fail again. I’m still none the wiser on what is causing issues, if any
 
pc-guy said:
A system can boot doesn’t mean it is error free. What you have done all morning hasn’t ruled anything out. So basically a total waste of time. Without methodologically testing each component you can’t draw any conclusions.

memtest only ever need to be done once and it take 4-7hrs to complete. You don’t even have to watch it as you can leave it going over night. It baffles me to this date that so many people expect OC on CPU and XMP and don’t actually test if the things work in the first place.

sigh...Good luck.
Click to expand...


I think my comment came across wrong. I had some time this morning, only up to an hour and it was very quick with what I did on the swapping. I will be testing each of the dimms and sticks as you originally suggested and will come back once with my findings
 
So I managed to run the memtest on each DIMM and had 0 errors across 5 tests, with doing an extra test for the other stick

I am noticing that I can add some oc to ram or cpu but when there’s a restart or I shut down and boot up about 10 minutes later I get the same boot failure detected. I’m at a bit of a loss with it. I might look to purchase the same RAM I have and test that to completely rule it out.

I’ve had the cpu cooler off and reseated my cpu. No bent pins or anything of concern. I feel it could be a mobo issue
 
This was with memtest86 from a usb drive.

sorry, I meant the reboot is if I restart from windows or shut down then use my computer later or as it’s booting back after I restart it seems to get stuck attempting to check the memory. The dram led stats on, flicks to the cpu then back to dram.

do you think it’s worth taking my NVME and using my original SATA SSD or am I overthinking?

Also, I’ve noticed that I used to be able to have a 4.8ghz overclock on my cpu at a 1.275v VCORE. I’ve noticed I’m having to bump that to 1.280 if I want it to boot.
edit* It’s getting boot failure at this voltage also. I’m going to remove it to not cause any issues to my cpu


I’m swaying more and more to motherboard now.

thanks for all of your help btw
 
Last edited:
pc-guy said:
dont OC your CPU. just run it as stock AUTO settings. leave eveything stock and run the RAM test with the stock then XMP. if stock passes and XMP fails or doesnt boot then RAM is the problem. you cant be OCing every component of your system and do diagnositic on ram. you just dont know which part of your system is failing that way.
Click to expand...

All memtest tests were completed with XMP alone, the DIMM slots do seem to take a while to boot when the memory is first added to them
 
pc-guy said:
it take time for the CPU to train memory and usually need a few reboots to settle down. it is completely normal. unless it stops posting then that indicate a problem. did you test all the slots at stock ram setting and it all works?

did you also do a clear cmos and load optimised default before commencing your ram testing?
Click to expand...

ahh I didn’t realise about it taking time to train. Yeah all tests completed at stock settings individually CMOS wasn’t cleared between runs but load optimised defaults was used.

I have just attempted (with no CPU OC) to put 1 RAM stick into each slot and add just XMP without any extra voltage and I was unable to get XMP to post once in 8 tests. Would this help to narrow it down?

If I’ve been able to get XMP applied, it normally gets stopped on the next shutdown or restart
 
I always assumed XMP dealt with the voltage.
When I attempt XMP it will attempt to boot for a while and I see the XMP LED on my mobo to come on and then go off, fans spin up quite a lot. Then it will boot and the logo comes up and it goes to the ‘Boot failure detected’ screen and have 3 options of load optimised defaults and boot, load optimised defaults and restart and enter bios. The odd occasion XMP will work after waiting 5 minutes plus for it to boot, but once I restart my pc or shut down and boot back up it will get caught by the spinning up for a few minutes and the. Majority of the time will show the boot failure detected error
 
I gave it a quick try with adding XMP and manually adding 1.35v, I even increased it to 1.40v+ but still no boot. I’ve seen some ram that’s 3200mhz and I’ll try it, I will keep it regardless of it fixing it or not and then get gigabyte to RMA this board for me.

I feel like it’s RAM now, I’m on default bios settings with a bit of added extra to VCCIO and VCCSA to test if it’s something on the CPU and I’m getting perfect boots and restarts.

Thank you for your help and guidance, I feel you’ve taught me a lot with over clocking from what I already knew
 
pc-guy said:
Good luck with RMa process.
Click to expand...

So on Friday my new RAM arrived. It was some HyperX 3200mhz which runs at 2400mhz standard.

I originally had issues applying XMP to the new RAM which was worrying but after manually adding VCCIO and VCCSA (both lower than 1.14v) it began to become more stable. So I decided to manually key in XMP settings and even use a dram voltage of 1.34v and all weekend it has been stable, the only time I’ve had boot failure detected is if I dropped the voltages too low, but I’m happy as I have a lot of wiggle room in regards to where the numbers are, I’ve even managed to tighten the timings for XMP.

I have done reading on other people using the Aorus boards with 9th gen i7 CPU’s, who have also reported strange behaviour with RAM, over locking and XMP.
 
You must log in or register to reply here.
Back
Top Bottom