Jump to content
Search In
  • More options...
Find results that contain...
Find results in...

CPU running 90-100+

I've got an Arctic Freezer II 240 (on an i9 9900k) and it had an Icy Diamond pad on it, until I noticed recently it's been hitting mid 90's on most cores, and 100+ on core 4.  I went back to paste today, and spread it over the IHS in as thin a layer as I could get away with and still have it on the IHS (any thinner and the credit card would be removing the paste lol) (AS5 because it was what I had from previously), and it's doing the same thing.  I went back into the bios, set everything back to how this video set it up: 

I have gotten a stable OC of 5.0GHz at 1.28 Volts and can't go any lower without stability issues.  But even then the temps persist into the same territories.  It's not until I throttle it all the way down to 3.6GHz that I start to see any lowering of temps.  The thumb... nuts?  (for lack of better words, it's not a screw so lol) are snug and just a little past but not so hard tightened as to cause damage to the threads or IHS.  I'm at a loss for words here.  I know that the i9 9900k runs hot and is "ok to run" at 90's+ under AVX loads but, I'd rather get back to the 80's I was seeing when I first put this in my system a couple of years ago.  I know that evap can happen, but I want to doubt that that happens in 2 years to this level, plus it'd be making air-bubble noises if it had gotten that low.  In fact I don't hear it run at all, but the fans do.  That's the part that has me wondering, I mean it never made noise to begin with but had good temps... however I don't know of any way to know that it is running.

 

I do notice the air is moving, but it's not pulling hot air off the radiator like it should be for heat of this level... ((edit:  Strike that last part, after letting it run for more than 20 seconds it is moving warm air off of the rad.  I just had to let the liquid saturate some  DOH)) Is there any way to actually tell that the pump is running besides sound?  I have the lead coming off of it hooked to the CPU_Fan header on my ASUS Maximus Hero XI WiFi.

 

It's also mounted with the Rad at the top of the case, fans under it pulling the air through to pull cool air in through the fins of the rad.

 

Oh, and it idles between 35 and 45, mostly around 42/43.

Link to comment
Share on other sites

Link to post
Share on other sites

If the pump is running and there's no clogs the tubes/pipes should be roughly the same temperature. If there's a cold spot in the tubes or on the radiator - it's clogged.

I am not allowed to have my PC specs in my signature.

Link to comment
Share on other sites

Link to post
Share on other sites

It's not clogged then as the tubes are all a neutral temp if a bit warm.

 

I'm beginning to wonder if I have put too little on the spreading out of it.

Link to comment
Share on other sites

Link to post
Share on other sites

Just a bit more information now.  It's running ok (60's or less) when in anything non-avx.  It's on OCCT or Cinebench that it goes this high.  Games run fine, so I'm ok for now, but I'm just curious on what this could be?.  If the pump is good, and fans are turning this shouldn't be an issue with this particular AIO, I mean I've had it at 5GHz and 85 or less with something half this size (a corsair h60 first gen I think) so I'm at a loss for what it could be without repasteing and trying again, but if it's transferring heat enough to keep it 60 or less during games, then that suggests that the paste is fine.  I'm just at a loss atm.  I mean I have ordered some KPX, but that won't be here for about a week so and won't help THAT much, so yea... at a loss...  any more ideas to try?  I am thankful for the ideas so far though.  I mean, what could be doing this such that AVX loads push it that much harder and into 110+*C territory?  hmmm

Link to comment
Share on other sites

Link to post
Share on other sites

3 minutes ago, OutsideOctaves said:

Just a bit more information now.  It's running ok (60's or less) when in anything non-avx.  It's on OCCT or Cinebench that it goes this high.  Games run fine it so I'm ok for now, but I'm just curious on what this could be?.  If the pump is good, and fans are turning this shouldn't be an issue with this particular AIO, I mean I've had it 5GHz and 85 or less with something half this size (a corsair h60 first gen I think) so I'm at a loss for what it could be without repasteing and trying again, but if it's transferring heat enough to keep it 60 or less during games, then that suggests that the paste is fine.  I'm just at a loss atm.  I mean I have ordered some KPX, but that won't be here for about a week so and won't help THAT much, so yea... at a loss...  any more ideas to try?  I am thankful for the ideas so far though.

That's why the AVX offset exsists

 

My 10900k for an example will hit 90-95 in cinebench.  But in games doesn't even brush 60

 With all the Trolls, Try Hards, Noobs and Weirdos around here you'd think i'd find SOMEWHERE to fit in!

Link to comment
Share on other sites

Link to post
Share on other sites

Thank you, but I've already tried that.  It wasn't doing this under AVX loads before hand, so I'm currious as to what could be done to achieve that same 5Ghz all core at 85C or less again with the same equipment if it is?  I mean I'm running less voltage now, it was 1.320 I believe now it's 1.28 (lowest stable with everything in that video set as told).  I even tried offsets that went beyond 4... even taking it back to stock clocks and it was still pretty hot.  So I'm wondering at this point if something went wrong or what.  hmmm...

Link to comment
Share on other sites

Link to post
Share on other sites

Well, to make things even stranger.  I went into XTU and played with my voltages again and got it down to 1.200 in one run, but then I tried lower and it went BSOD.  Then on the next run of windows it would not stay stable anywhere near there.  So I went back up to 1.280 eventually, and that is where things stayed stable.  Then I went to my Cashe and turned it down to 42 and found it didn't increase stability so back to 45.  Then I started playing with offsets again.  I wanted to see if it was going to do the same as before, and force me all the way down to 3.6Ghz again with minimum temp. differences, or if I'd get a different result.

 

Strangely, around an AVX offset of 9 I was getting temperature reductions to under 100 again.  I set it to 10 and I was consistently 96C or less.  For the moment it seems an offset of 9 is the "butter zone" of not hitting 100+C but allowing at least above 4.0GHz for AVX instructions, if only by 100MHz.  Now the question is, why is it being like this 2 years in.

 

I watched the temps, and it's cooling off mega fast after the test is concluded, so the pump seems to be working on that front.  (I mean, given that I'm running the fans and AIO at 100%, but still...) So it's just saturating quickly and stabilizing by around the end of the test (I could run it longer and of course it would continue a bit of a rise but I'm just going on single runs for the moment).  Hmmm, could it be that the Arctic Freezer II (Mark 1?) just isn't up to the task long term for one reason or another?  But strange still that last time I did this battery of tests I couldn't get the temps to come down at all really till I was running all core 3.6Ghz, and then this time I could get temps to actually come down a good bit overall at an offset of 9 to 10.

 

I mean I did blow it out again, and not much dust came off of it at all (it's been a good few days now).  However when I took a fan off, it was as clean as it could be so yea not much dust was there to begin with today, and the tiny bit that came out wouldn't cause it to hold THAT much temperature.  I mean I know dust can do that, but usually it's a good layer of it that causes that, this was like perhaps just those few days worth of dust that had re-collected on it.  So I doubt it was blowing it out again that caused it to suddenly start cooling things under AVX load... strange indeed.  At least I was able to get some degrees off of it at max.  While the other cores that weren't maxing out in temp were substantially lower doing anything from 50's to 80's under load so... strange indeed.  More to chew over.

Link to comment
Share on other sites

Link to post
Share on other sites

OK, To close off this topic:  I FIGURED IT OUT... OMG I'm such a ... Linus.  ...

 

No, I didn't drop anything (except my temps... *burn!*).  But The issue seems to be my IHS... and my super thin application of thermal paste this last time.

 

You see, this time not only did I go for the Kingpin KPX (finally got in in after a week of waiting from amazon seller that had high ratings for their speed... :rolleyes:  , but I also went and after spreading it thin like last time I put an extra (tiny, less than a quarter of a pea) on the center... and after some faffing about with the nuts falling into the pc again, I got it tightened down... I even went past "just past tight" to push it as closed as it could be so to speak.  Here I went to the bios and checked that I was in 5.0 with 0 AVX offset... and forgot something here but that's for in a min.  I saved and reset.  After closing everything that loads up with windows for me and getting some background stuff I'll need to uninstall out of the way, I ran HWMonitor and Cinebench r23... and got my highest scores ever and my lowest temps ever while under load on both Multi and Single Threaded runs!  I even got my max temp lowered to 82 ON 2ND RUN!

 

Yea, I don't know what happened over time to the thermal pad I had on it, but even though it worked great originally it was not working anymore... then my first attempt at pasting this system in years... when I pulled off the cold-plate today, it BARELY had any contact... I mean even made me wonder how in this last week I could even game without it maxing out the temps, it was THAT bad... thank god I pulled it off today.

 

On repasting it with KPX today instead of AS5... I got the coldest temps ever and found out my IHS is concave...

It was that simple.  Now how my pad of icy diamond worked originally but lost out over time is a mystery for me.  But I have a working layer of paste now, and I know my cooler is working fully still!  yay.  Thanks for any ideas and help that anyone gave, despite it being as simple as repasting...

 

Oh, and that setting I forgot to check... the Voltage was not set at my usual "this is my stable place" setting of 1.320... I had it at 1.280... which wasn't a stable place to be, it was just a test from earlier this week when I was still troubleshooting things.  And yet it runs perfectly stable at this with everything else set as was.  Yea, I know that this is part of why it's running cooler than ever but still, kpx has allowed me somehow to get a stable undervolt of my usual stable place... that was just at stable with my pad and with any paste before.  Yet now 1.280... makes me wonder if I've got a chip that will go lower and still stay at 5.0 or perhaps push harder to 5.1 or 5.2... or such... but that's for a later date.  Thanks again.

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share


×