Jump to content
Corsair Community

H100i GTX hitting 105c package temps


sotti

Recommended Posts

I doubt the NT-H1 went bad it's in a syringe always stored with the cap on.

 

I just ordered a scythe mugen 4 and some artic MX-4 paste off amazon to do testing with.

 

No, I wouldn't bet on the TIM going bad, but we are running out of possibilities and down to what's left. Hopefully the next test will provide some answers.

Link to comment
Share on other sites

I'm going to have to give it some time, but it seems the air cooler is at least as good as the h100i for me, perhaps even 4-5c cooler.

 

But like I mentioned every time I do fresh remount, it seems to run cooler for a bit.

 

So stock speeds I'm load at 55c and idle at 37c, ambient is close to 29c compared to tuesday when it was only 25c, so that seems to point to a push between the mugen4 and the h100i. As previous with everything stock adding core voltage to 1.25v, the core temps jump to right around 70c with load. If adjusted for lower ambient, then it appears we might be slightly cooler with the air cooler.

 

 

The factory overclock to 4.0 does show a bit of difference in favor of the water cooler. My idle is closer to 45c instead of 38c, which is more than simply ambient can explain. OTOH load temps were similar in the 70-75c range.

 

I can't really reach any conclusion, I was hoping the water would be a good 7-8 cooler, which being a hotter night testing the air cooler should really favor the H100i, but I'm not seeing that. Conversely, it's not as if the air cooler separates itself enough to make me think that the h100i isn't doing close the same work.

 

Bottom line it's really disappointing, but I'm not sure it points to a defect. I'm curious as to what your opinion is.

Link to comment
Share on other sites

The thing that really bothers me is the change in idle between various overclock levels. The peak values are so variable and there are so many factors involved, but in the two years I had been submerged in this X99 platform, I can't remember anyone having idle temps in the mid 40's, unless they set their processor to 100% with a corresponding high Vcore. The large jump between stock and mild overclock makes me thing we are missing some setting or trick, but I don't know the ASRock BIOS enough to shed light on the issue. I wonder if it would be worthwhile poking around the ASRock forums (assuming they have one) to see if there is something I am missing.

 

I am going to send you a few screens of my old 5820K on a Dark Rock Pro 3 air tower. Not all my records survived the purge to from Win 7 to 10, but I have a few left. Comparing other AIO water coolers is fairly easy. Most have similar density radiators, fan speed is less critical, and flow rates similar or inconsequential. Air towers are much harder. Lots of variation is radiator design, lower densities, and more dependence on fans speed and design. My DR Pro 3 ran about +10C over my H110 at load. I don't think you can straight apply that your situation, but I will send some data.

Link to comment
Share on other sites

I guess another question is how fast water temperature should fall back to normal after you kill the process load.

 

In my mind everything looks normal, the water gets hot (jumping 5-10c in maybe 3 minutes of 100% load), then the water cools off (close to pure idle temps after 5 minutes of rest). But the absolute amount of heat is just way off the charts compared to what I'd expect from a stock 5820k.

Link to comment
Share on other sites

I was going to knock my setting back to stock, but then decided to leave them to measure a higher voltage, rather than my unusually low CPU defaults. 4.5@1.275v. Cache 4.0@1.17. (load Vcore 1.29v) I kept my radiator fans at the bare minimum (675 rpm) to de-emphasize any difference between the 240mm and 280mm radiators.

 

OCCT 4.2.2

 

23C ambient

 

Idle water temp 25C, cores 25-28C

 

Instant Load to 55-60C. 1 min load 65-70C. (of note: OCCT does a 1 min 'warm up' before really kicking in, even when you don't specify a warm up period. There will always be a jump at 59 seconds). At 5 minutes I had climbed into the mid 70's, but this is a result of the low fan speed and rising water temperature.

 

Water temp went up from 25 to 32C in the one minute and further after that. However, since I crippled the fans to put more emphasis on the thermal transfer ability of the cold plate, I would not consider that a 'normal rise'. When I re-ran the test with a fixed 1300 rpm, the whole curve shifted down 3C and water temps returned to normal within 20 seconds. 29C max water temp. 5 minute mark gave the same results as at 1 min. No further rise in water temp or core temps with better fan speed. Mid-60's over that duration.

 

I'm afraid all of this suggests something at the CPU end of the chain. Whether that is internal to the chip, TIM, cold plate, or some sort of physical contact anomaly, I don't know.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...