I'm still having issues...


After almost 2 months with this motherboard and spending time on the forums, internet search, Discord, etc... I've come to the conclusion that something is wrong with this board or bios.

The primary problem since the newest BIOS 1.14 is that NOTHING I change for the CPU in terms of voltage, under or over, OC Robot, ANY OC whatsoever, nets a code 71.  OC Robot actually worked before the BIOS update.  I have pulled the waterblock (Heatkiller V Pro) adjusted the contact frame several times, etc. and nothing works at all unless the CPU is at FULL AUTO.  I can't even change the voltage to adaptive and leave it there in the BIOS.

However, until today, I was able to adjust everything that ELEET software allows and am able to achieve a very respectable and stable daily setting 5.2 all P-Core and 4.2 all E-Care and x40 Ring. With either adaptive or a 1.3 override on the Vcore Target Voltage. Someone from Intel told me to leave it on adaptive with all core OC.  But now, ELEET has started acting up by not keeping the BCLK at 100.  When I Load any profile, it hangs for over a minute and fails with an error because it's trying to load the BCLK and PEG/DMI at 50.



I thought that by buying the KP board I would have the ability to set a respectable OC in the BIOS as I have for all of my rigs.

Here's a pic of the rig for reference.

I'm not even sure EVGA cares much about their motherboards the same way they do/did their GPU products. After this announcement from the CEO, we're going to have to see what happens.  AMD may just have been given the luckiest break in the history of consumer technology.

I have a ticket in with the EVGA support team.  I'll report here what comes of it...  For now, ELEET is working again.  It's C23 stable at 5.2 all P/4.1 all E with 40 Ring.  RAM is clocked down to 6200 only because I had to reset it in the bios because of my CPU testing.  



But it runs stable across the board at 6600 and 6800.  But at 6800, it gets warmer than I like with no noticeable performance boost outside of bench mark scores.  The motherboard does REALLY well with RAM OC in terms of primary AND sub-timings but I am running Hynix M-die so that helps.  I'm just hoping for the same result when DDR5 performance sticks hit the market with Hynix A-die.  The next gen DDR5 will run at 8000MT... insane AND that will require more from the CPU which is where I'm concerned for 12th gen.  



I agree that 13th gen may solve some problems, but man... EVGA needs to step up for us on that in terms of the BIOS they give us.  This is Classy and KP!!  Where's the love for reputation, it's feeling like they are just giving up with these newer product outcomes outside of the FTW3 GPU and power supplies which have been STELLAR to date.   I could say so much more... but I digress.


More later...

I'm currently getting this about every time I turn the computer on. I then go through a couple times of trying to hit the reset button or power cycling the power supply switch, but eventually it boots up and seems to work perfectly fine once it's booted.

zippytek
[snip]
man i'm sorry to hear you're still having this issue. have you tried reaching out to EVGA support to see if this might be a sign of failing board? i've been through 5 different 12th gen cpus on z690 dark kingpin and have never see a 71. i know it's a huge pain for you since this is your daily, but seeing all you've tried with similar results, i may explore the RMA at this point. i think they offer cross ship, so maybe that's an option to minimize downtime?
 
wish i had more ideas on what to try. 

Yeah, I'm thinking the same way. I just don't want to deal with it right now and as long as I can do some combination of rebooting and get it back, I'll keep plugging along for a little while. That said, I'm really unsure of what EVGA's inventory of motherboards looks like and I may not want to wait too long lol. I was planning to wait at least long enough to see independent reviews of the 13900k so that if I wanted to go that route, I can do it all at once.
 
I'm really not sure how much of my issues have just been chasing this board around and every time I think I have it figured out it seems to come back eventually. It either works perfectly fine and everything is wonderful, or I can't boot. Very strange. Also, I don't know that the g6900 really tests things as it has a fraction of the features and clock speed of the 12900k. I'm kind of wondering about that with the 13900k speed rumors, as I'm wondering if one of those would even boot at normal speeds if this is a clock-speed-triggered situation. At this point, I have no idea really. I do have mixed feelings seeing b3l13v3r have the issue too...at least I'm not alone haha, but I'm not happy someone else is going through this either.
 
In any case, I have gone full circle to think it's probably not the CPU itself. I do still really appreciate the time you spent trying to help me troubleshoot this, so thank you for that.
 
Also, b3l13v3r, The bclk issue is a bug that others have reported on here (I've seen it too). It was also present on z590, so it's not even motherboard specific, it's just an ELEET bug.

B0baganoosh
 
In any case, I have gone full circle to think it's probably not the CPU itself. I do still really appreciate the time you spent trying to help me troubleshoot this, so thank you for that.

dude, no worries at all. even now i'd be open to help you with it if you think the facts lead to it being the cpu. i honestly don't think that's the issue and as we discussed before it was a good enough bin plus avx-512 enabled that it wouldn't be hard to resell if you're going to upgrade to raptor. for avx-512 enabled benches, the oc community can't touch 12900K even w the fastest KS since it is fused off, so that should at least give you a silver lining should you need to switch boards and/or choose to upgrade to 13th gen in the coming months.
 
https://hwbot.org/benchmark/y-cruncher_-_pi-1b/rankings#start=0#interval=20
 
for 16x, KS doesn't even break in to top 10 bc of avx-512 fused off

b3l13v3r
Soooooo.... sh*t!!  This just fixed the issue I was having with booting the system after making ANY change to the CPU settings. And I mean... ANY!!

Once I changed C-State Enable then Enhanced C1 Disable, I then went to change the OC to what was stable using ELEET which is 5.2 all P's / 4.1 all E's / Vcore set to 1.3 override and it booted!  I tried setting the Ring clock to 40 in the BIOS as well but that would only cause a post code 71 again.  I can set the RING/LLC up to 42 in ELEET and it will stick until a reboot, but I cannot change that in the BIOS and get past post code 71.  I will try more later, but I'm STOKED to have found this one post that I somehow missed while troubleshooting over the last month.  I feel stupid for not finding and trying this.  Hopefully it sticks for good and no reflash, or worse, RMA.

I'll report any failures or changes.  I see that 1.15 is out and I may test that on one of the other bios channels when I have time.

I remember that post, but it didn't seem related to the problem I was having so I didn't change the setting before. I did just change it and I didn't get POST code 71 when I booted...but I did get POST code 62. 
 
So 71 is "South Bridge DXE SMM initialization is started"
62 is "Installation of the South Bridge Runtime Services"

Either way, seems to be related to South Bridge initialization. Anyway, similar to the last couple weeks, I can hit the reset button and the second time it boots up normally.

I wanted to update this thread as it was a bit of a wild ride and it has been months. After my last post, I had been using the factory frame, the EK backplate, and just a lighter tightening of my water-block. I hadn't really noticed any thermal degradation, but it was never quite as good as it had been that one short stretch I had the contact frame working. I did occasionally have a code 62 or 71 on a cold boot, but one reboot and it was working fine. I had sort of assumed that if it ever got worse, I'd have to RMA, but it hadn't gotten worse.
 
I revisited this after buying a 13900k, as I noticed the 12900k had started to develop that high-spot-low-spot look to the paste when I took the block off for the first time in a few months, so I figured it was worth putting all my experience from that previous drama to use and take a fresh stab at it.
 
I think I can try to summarize where I went wrong and why knowing what I know now I was comfortable trying it again with the new 13900k (maybe I should put this in that thread and the contact frame thread lol):
 
1. My Optimus water block doesn't utilize a backplate by default.
2. I did the install with the case in the upright position. So without a backplate behind the LGA1700 backplate, the LGA1700 backplate can come right off when you take off the factory bracket. I was careful, but I think there were times where this moved out of position.
3. I was also tightening my water block too much I think. There are no springs or anything, it's just knurled nuts on posts that you finger tighten and maybe I took their "just tighten until you can't finger tighten anymore" too literally.
>>Combine these items and what I believe was causing my issue was flexing of the board around the socket and all this flexing and potentially misaligned brackets led to intermittent contact quality, which led to issues.
 
So my adjusted successful steps:
A. I got one of the EKWB backplates they offered for free (you just pay shipping basically) because their old one wasn't compatible with LGA1700 and they made a new one to offer compatibility for their coolers. I then bought bolts that thread through the back of that EK backplate, some nylon washers, and some lock-nuts for it. The bolts are the same thread as the Optimus mounting posts (the block-side-size, not the original motherboard-side-size).
B. I put the original z690 bracket back on very carefully and made sure that it was tightened down in a manner that ensured the LGA1700 backplate was once-again correctly and evenly mounted.
C. Then I very carefully mounted the EK backplate behind that into the motherboard with the bolts, putting the nylon washers and nuts on the front of the motherboard. It is also important not to over-tighten this...I have a habit of that I guess(classic tighten until the bolt breaks, then back it off a quarter turn).
D. So with that in place, I can take off the factory bracket without the factory LGA1700 backplate moving as the EK one holds that in place now. This is critical.
E. This time, when I followed the steps from Thermal Grizzly and GN on installing the aftermarket bracket, it went much smoother.
F. I then put my water block on and adjusted their instructions to "just tighten until the nuts no longer easily finger tighten anymore" (it's subtle I guess lol) and I haven't had any of the frame-weirdness with the 13900k at all.

frankd3

B0baganoosh
I wanted to update this thread as it was a bit of a wild ride and it has been months.
 
(good info snipped)
 

I, for one, am glad you posted this followup.
That finger tightening you ended up with is exactly what I learned with my EK AIO (on the 11900k). When I did that I finally got the cooling performance I expected.
I like the way EK gives me long easily reachable thumb nuts so I could do that with ease. I'll never use a screwdriver to tighten down a cold plate again.
Learned that lesson the hard way!
 
So, you're happy with the contact frame and you're using it on the 13900k?
It's funny that I haven't heard anybody mention using the contact frame on 13th gen, only 12th gen. Same design and dimensions and same ILM.

Yeah, the Optimus thumb screws also don't have any slots or tool-implements to make sure people don't throw a screwdriver or hex key on them as well, but I still seem to have over-finger-tightened them. Also, I always cranked it down on my 11900k and never had a problem. 
 
Yes, I'm so far quite happy with the contact frame on the 13th gen. I've seen several reports of other people using them, but not a ton here specifically mentioning it. 

zippytek
nice! glad to hear you figured it out. i have both TR and TG frames at home and I never used either bc my dark never had mount issues and frankly I wasn't convinced I would install it without causing memory oc issues.
 
i've also seen Asus specifically say not to use a contact frame for raptor lake boards (z790) so I guess we'll see if these issues repeat in the coming months! i know several folks using a frame on their z790 Apex without issues. and then there's the issue of dead boards that now plagues them (usually one dead mem slot)
 
enjoy!

Oh boy!
Thanks for the info on the contact frame.
 
By not having mount issues are you judging by temps alone or have you seen the paste spread pattern?
 
EDIT: a little googling and I found that the stock ASUS Z790 ILM screws are not long enough. So people who want to use the contact frames have to get their own screws to replace the stock ones. This is specifically mentioned with the Thermal Grizzly frame.

zippytek

frankd3

zippytek
nice! glad to hear you figured it out. i have both TR and TG frames at home and I never used either bc my dark never had mount issues and frankly I wasn't convinced I would install it without causing memory oc issues.
 
i've also seen Asus specifically say not to use a contact frame for raptor lake boards (z790) so I guess we'll see if these issues repeat in the coming months! i know several folks using a frame on their z790 Apex without issues. and then there's the issue of dead boards that now plagues them (usually one dead mem slot)
 
enjoy!

Oh boy!
Thanks for the info on the contact frame.
 
By not having mount issues are you judging by temps alone or have you seen the paste spread pattern?
 
EDIT: a little googling and I found that the stock ASUS Z790 ILM screws are not long enough. So people who want to use the contact frames have to get their own screws to replace the stock ones. This is specifically mentioned with the Thermal Grizzly frame.

info came from safedisk, one of the top Asus sponsored overclockers. if it was easier to share photos here i could show some of my mounts from the past year with the dark, but yes paste spread was never an issue for me on z690 dark like it was for other boards. i always use a heatkiller block with 1700 backplate and mount kit so i'm not sure if this helps enough to avoid the bend or what. i use the same backplate on z690 apex and while it's still fine, it isn't as nicely spread as dark.
 
this being said, for my needs i'm usually using a liquid nitrogen container which is *much* heavier than a water block or aio, so it is entirely possible this helped to "break in" my ILM since I ran more sessions with a heavy copper pot on the board than water block. just my experience, and i understand it is not that helpful for daily users. 
 
i haven't run the apex cold yet, but plan to in the coming days. spread with waterblock and stock ILM is manageable enough but i run well up in to the 90s for some benches with 60 P core, 55 ring just under 1.5vcore so it's to be expected. 

Thank for explaining.