SOLVED Which of these temps would you believe?

[Neostarwcc]

Okay so I overclocked my i5 3750k to 4,000 mhz, 1.2v and the BIOS says that the idle temps are between 50-60c. I figure this is okay and I probably won't max the CPU on a stress test.

So, I do a stress test and monitor it on both Core Temp and SpeedFan and I get the following:

According to CoreTemp, I maxed the processor yet according to SpeedFan, I pretty much achieved the idle temps that were reading in my bios? I'm not sure at this point, which to believe. I don't want to max my processor under ANY circumstances so I want to make sure that it isn't happening. It also says in SpeedFan that my idle temps are 38-40c and in core temp it says they are 48-50c.

What makes me believe SpeedFan more is that my GPU temps are correct. My GPU always runs 35-40 idle and 60-65 on load. But CoreTemp has never failed me before...

I am using an ASROCK z77 Extreme 4 for a mobo by the way and am running the processor and GPU on water which is why I figured it went in the water section.

 

[iLoki]

You can probably do better than 4Ghz on 1.2v, and those temps (according to CoreTemp) seem WAY off. Under water, you should not be idling at 50-60°C, especially not at 1.2v.

Even before I delidded, my idle temps were 36-40°C at 1.25v. Under heavy stress (LinX/IBT) I would see a max of 88°C

Leads me to believe that CoreTemp is having an issue properly reading your temps here..
Can you check the temps with something else? HWMonitor, RealTemp, OCCT, etc?

Also can you provide some info about your loop? How much rad, what CPU block, what GPU block, what fan config are you running on the rad(s) (push,pull,push/pull?)

 

[Neostarwcc]

That's what I figured, because my main desktop doesn't go over 60c period. And I can probably push my 4930k even higher and not have a problem with it at all.

At first I figured it was the water since the water was in fact old, well I changed the water and it said there was a 5c difference in coretemp/the bios. It may be due to ****ty paste or paste placement on my part too. I was in a rush to get this built and going the other day and I was using kinda ****ty paste (the paste XSPC gives you with their kits).

I should be using arctic silver 5 or some other premium compound but I didn't think it would make much of a difference and I don't have any left anyway I wasted the rest of it in December when I built this computer.

But, anyway, I don't believe the BIOS at all because there's no way I should be idling on 60-70c on water or hell, even air. It's not like I pushed the processor anymore than stock and even stock apparently is 60c. Things... just don't add up I might go with another monitoring program and seeing what they say it is.

 

[Neostarwcc]

I'm using a raystorm ax240 kit and using the loop that they say to use. I think I can find the manual somewhere on google but I'm using the loop on that one with the radiator on top and the pump in the front. As for GPU block I'm using this one here:

http://www.amazon.com/EK-Radeon-Lightning-Water-Block/dp/B0096FCN8M

I don't have a camera or I'd just take photos for you.


Anyway, trying different programs now.

 

[iLoki]

A single 240 is pushing it for a CPU+GPU loop, but your temps should still be better than that.

My setup is actually very similar to yours, I'm using a Raystorm cpu block, and the EK full cover r9 290X block, though I have an additional AX360 in the loop.

If BIOS is claiming 60-70°C on idle, then you very likely have a bad mount. Either way too much TIM or not enough contact between the block and the CPU. The Raystorm has 2 sets of adjustments on it, there's the standard screw (the top thumb screw) and then the thumbscrew underneath that needs to be tightened to ensure good pressure against the CPU. Try tightening that bottom set of screws a bit to get some better pressure on the CPU and see if that helps at all.

 

[Neostarwcc]

I agree that it's pushing it but my friends said that I could do this on a kit so I did. I didn't care about hard drive or ram temps/overclocking so I figured this was the cheapest and best solution. I'm doing it in my main desktop too although adding additional loops.

One interesting thing to point out is that it only stayed at 4,000 mhz for about 3 seconds and then to save the processor had to lower it to 1,600 mhz...

Something isn't working right.

Okay, it may well be the mounting because OCCT is giving me the same results.

Core #0 is what speed fan was reporting as "normal" this stayed under 50c during the entire 1 minute 30s test
Core #1-3 are almost all maxing out. and max out within 6 seconds after test start.

It's funny how Core Temp read Core #0 as being over 100c too but regardless this tells me that it's a bad mount of one core is fine and the others are maxing out.

So, time to check my mounting and make sure I mounted it correctly. I hate that part because it usually takes me almost 3 hours to mount it alone, because you have to try to hold the backplate, get the screw through the hole, turn it and try to get the other fastenings in there. Pain in the ***... oh well. If I did it wrong, I did it wrong.

 

[RGE]

bad/loose mount or left plastic on bottom of your waterblock. your cpu is throttling b/c core temps are at tjmax near 105C. speedfan isnt reading any of cores.

 

[Neostarwcc]

I've remounted it and the issue still exists, I'll buy a tube of arctic silver 5 it's only like 5 bucks. Could it be that I've received an unlucky processor? I had a 3770k in there before and had no temp problems at all. I downgraded to a 3570k because I wanted to save money. $150 vs $300.

Meh, I don't know I'll grab some paste anyway I should be getting some since I'm out of the good stuff anyway.

Thanks for your help I'll continue to investigate this issue.

 

[m0r7if3r]

Is the rad warm? I'm guessing dead pump if it's not a bad mount.

 

[Supertrucker]

Your gpu temp is also quite high. My 290 loads in the low 40s with a heavy overclock and +100mv. Granted, I have a lot more radiator but 65° on a gpu is air cooling temps. Even the aio 120.1 cooler with the new nzxt bracket was getting 45° load in reviews.

It seems to me like you have an almost dead pump, I say almost because gpu blocks handle low flow better than cpu blocks and if you had no flow your gpu would throttle under load too.

 

[GTXJackBauer]

At first I figured it was the water since the water was in fact old, well I changed the water and it said there was a 5c difference in coretemp/the bios. It may be due to ****ty paste or paste placement on my part too. I was in a rush to get this built and going the other day and I was using kinda ****ty paste (the paste XSPC gives you with their kits).

This makes me think you might have build up in the loop possibly. You state the water was old and after a remount, the temps barely changed. You say you rushed the build but did you do a complete clean out of all the new parts for debris etc before assembling?

Is the pump alive? Rad, tubes, both blocks warm as well as the air exhaust from the rads? Could be a dead pump as mor stated. I would also go with the premium TIM once the issue is resolved. I'd start investigating your cooling solution after you remounted it and the problem still exists.

 

[Neostarwcc]

It was the pump, I had the pump replaced and it's working fine now. Closing topic, thanks for your help guys.

 

[GTXJackBauer]

It was the pump, I had the pump replaced and it's working fine now. Closing topic, thanks for your help guys.

Glad to hear it all got resolved and get a peace of mind.

 

[m0r7if3r]

Is the rad warm? I'm guessing dead pump if it's not a bad mount.

It was the pump, I had the pump replaced and it's working fine now. Closing topic, thanks for your help guys.

I love it when this happens...not when your pump dies, just when I manage to get it right on the first guess.

 

[GTXJackBauer]

I love it when this happens...not when your pump dies, just when I manage to get it right on the first guess.

.....................