Dabadger84

alpharius194
I think I just came the closest I’ve ever been to having a heart attack. After these posts I pulled up hwinfo64 to check my settings and realized my cpu had stopped going over base clock and wouldn’t draw over 0.92v. Fixed it by resetting the whole pc to default settings and starting from scratch... have you come across anything similar I’m trying to figure out why it happened. My only conclusion is the pc got pissed it was without any power for two days during my move but that doesn’t seem right.

I've had some weird behavior in reference to whether or not my GPU wants to actually output video that's been going on since I assembled the system, sometimes I have to literally unplug it (from the computer) & power cycle the monitor to get it to display that Windows logon screen is there.  Other than that & the not posting/training when I'm trying to test memory timings (which I think is normal... maybe? Maybe I'm just not hitting the Retry button enough times?) I haven't had an issues... but then again my system is on every day.  Tomorrow, assuming the Kingpin HydroCopper Block arrives on schedule, will be the first time it's not on for most of the day, and that will be of course because it'll be being operated on.
 
I haven't noticed my CPU doing anything like that yet.  Other than the FMax option I mentioned before resulting in lower boost clocks etc, I've had no issues with it getting stick or not performing properly.
 
 
I plugged in 3800MHz RAM (just voltage bump nothing else changed from DOCP) @ 1.5V, FCLK 1900MHz @ 1.125V and booted with the CPU still at 4.65GHz @ 1.275V LLC Level 4, and got some swanky numbers... pretty sure this is almost the highest R20 I've gotten.  I have no idea if it's stable or not, haven't done any prolonged testing yet:
 

 
Tempted to throw Time Spy at it just because I know that will result in a new high score.  Probably will shortly.

Clovis559
 
 
Those are some pretty dialed numbers! My 4.85 was only 12700ish. At 4.85 I cannot get the system to boot any higher then 3600 on the ram.
Any update on the sexy waterblock?

alpharius194
I legitimately have no idea what happened other than maybe I should’ve disassembled it to move rather than just pack it with bubble wrap. If it happens to anyone else, my solution was to use the bios to place everything back to default, boot into windows and then back to bios again to turn back on PBO, DOCP etc. going to turn my cpu down some from 47.5 down to 47 and 1.35v and leaving SoC at 1.15v while I troubleshoot this. The gpu was acting perfectly fine which is why I didn’t really notice it, I wasn’t using an FPS counter and the kingpin just powers through most cpu bottlenecks. The closest I’ve gotten to an answer was an issue with a 3600 non-x that got stuck at base clock and 1.1v due to a bios glitch but I’m really hoping I didn’t damage the cpu/motherboard moving. Even if I manually set the voltage the clock would stick to 37 and only budge to put cores to sleep.

If 5900x/5950x’s weren’t as hard to get I wouldn’t be sweating bullets... though it’s not like I want to spend $550 again. Hoping I get my block soon so I have an excuse to drain this case and give everything a once over. Badger post some assembly pics when it arrives so the rest of us can be jealous.

wodell757
Dabadger84 hope tomorrow you get your block and post pics and maybe a step by step?   Lol .  I did read the manual and looks like you have to take to back plate off so did you get extra thermal pads or just going to use the ones it comes with.
I did see they recommend very small screw driver 0 and 1 size so I might I have to order a set just in case.  And looking at the P360 kit from Ekwb I will have to wait as its out of stock.  
 
 

Dabadger84

alpharius194
Is it possible for you to run the loop in this order, Pump - GPU - Top rad - CPU - Front rad - pump? It's how I ran my old loop and it kept hoses pretty small. So when I moved my brother's pc I definitely took everything out. I fell behind packing everything else and when EVGA only released blocks to you and one other forum member (to our knowledge) I lost the motivation to drain it and disassemble it all. Everything seems to be holding up now so i may mark this one down as a bios issue because I'm on an older bios/AGESA. Waiting for bios 3501 to come off beta for my board before I update again. Ran R23 multi-core at 23900 which is pretty good for 47 on my cpu. The highest I saw at 47.5 was ~24100
Edit: I'd say it's running okay again. 21155 is my second highest timespy and my cpu is tuned down a little
https://www.3dmark.com/spy/19961867

The way I want to have the top radiator, short tubes aren't going to be a thing. lol  I want the fittings to be towards the front of the case so if they DO by some weird chance leak after install in to the system itself, it will be leaking on nothing, instead of components.
 
And I definitely want the GPU's heat to be being dissipated by the 480 if at all possible.
 
My GPU is acting very odd with Time Spy today, I can't get my graphics score up where it was on my 20279 run again, my CPU score is higher than the that run, but I can't get the GPU to stay in the 2060-2115 effective clock range and spit out a proper score for some reason.  Pretty much gave up on that since the block being installed tomorrow means I have to re-set/relearn my OCing on the GPU because I'll be hitting higher boost bins due to lower temperatures, at the same offsets, so only my undervolt/OC curves are still useable once the block is on.
 
I haven't run R23 yet I don't think.  But I did test 4.65GHz/1900 FCLK/3800 RAM with my usual burn in stress test for an hour & got no WHEA errors in HWInfo, it passed the test... Once I get to a point where I'm at settings I actually want to run for regular or 24/7 usage, I'll do a lot more stress testing with OCCT and whatnot, both on the RAM & the CPU, to make sure everything is at least as stable as it can be - the last thing I want is to be in the middle of something important & get a blue screen. lol