2500k 5ghz showing wrong temps?

Brackle

Brackle

Old Timer
2FA
Joined
Jun 19, 2003
Messages
9,101
Just curious what temps people are getting with a 4.8-5ghz overclock?

I recently just got my new cpu/mobo below. And after 24 hours of prime95, my CPU was showing 66c?

I mean my 5ghz is at 1.5v....I honestly expected it to be WAAAAY higher, specially since im on air cooling...

Thanks for your time.
 
I would expect his signature to have the correct cooler.

Cooling: TRUE-120cu lapped
 
Nice temps. Did you lap the 2500k?

Not sure what my Prime95 temps are at 5GHZ. Have you tried Intel Burn Test? 24 hours of prime seems extreme for stability testing.

What type of Thermal Paste are you using?
 
SonDa5 said:
Nice temps. Did you lap the 2500k?

Not sure what my Prime95 temps are at 5GHZ. Have you tried Intel Burn Test? 24 hours of prime seems extreme for stability testing.

What type of Thermal Paste are you using?
Click to expand...

No I did not lap the 2500k, only the TRUE120cu is lapped.

I am using Noctua NT-H1 thermalpaste.

I did a 12 hour prime95 while I was at work last night, left my the room closed. And still temps only got up to 67c.

So I am just going to assume that its normal.....I have read you dont want to go for 4.8-5ghz without water cooling....Well I guess this TRUE120cu is proving those comments wrong.
 
Is that a little high on the voltage? I run mine at 4600Mhz @ 1.310 volts, I can go higher to benchmark but i was afraid at 5ghz @ 1.40 volts was to high. My temps are good Idle 40C at those speeds. Whats the recommended voltage on these and whats safe? Maybe i can go a lot higher.
 
Marcdaddy said:
Is that a little high on the voltage? I run mine at 4600Mhz @ 1.310 volts, I can go higher to benchmark but i was afraid at 5ghz @ 1.40 volts was to high. My temps are good Idle 40C at those speeds. Whats the recommended voltage on these and whats safe? Maybe i can go a lot higher.
Click to expand...

No Idea.

I have heard so many people say that this voltage is right and this voltage is right. so TBH...

I guess the only way to find out is to test and see how far they can go.
 
Ya i was happy where i was at, I Guess ill crank up the voltage and report back with temps. Im off of work today, No kids, No wife just a Monday with my PC ;)
 
I don't use Prime95 becasue I don't think it's good enough. I'm on water (real water, not H50 crap) and my 2500k needs 1.475v to get there. With IBT it will touch 82c and hover 78c.
 
I would of thought that your 5.0Ghz Prime temps would have been higher?? What kind of fans are you running? What kind of ambient temp? My 2500k at 5.0Ghz under water "crunching" WCG (load) are running peaks of ~58/62 C. Prime would run even hotter. Why don't you post up a screenshot? Maybe you have a rather cool running chip?
 
I have heard so many people say that this voltage is right and this voltage is right. so TBH...
Click to expand...

I believe voltage needed would depend on how good your core is and also other components of your system like motherboard.
 
PolRoger90 said:
I would of thought that your 5.0Ghz Prime temps would have been higher?? What kind of fans are you running? What kind of ambient temp? My 2500k at 5.0Ghz under water "crunching" WCG (load) are running peaks of ~58/62 C. Prime would run even hotter. Why don't you post up a screenshot? Maybe you have a rather cool running chip?
Click to expand...

I will in a few, just got home and got the beast up and running.

I use these fans, but the 1600rpm versions

http://www.newegg.com/Product/Product.aspx?Item=N82E16835185065

Using an Antec 1200 case as well
 
5ghz.jpg


After 13 minutes.
 
FireBean said:
I don't use Prime95 becasue I don't think it's good enough. I'm on water (real water, not H50 crap) and my 2500k needs 1.475v to get there. With IBT it will touch 82c and hover 78c.
Click to expand...

You sir have made me a changed man. IBT really does stress the PC way better then prime95.

After 2 linpacks completed I am seeing temps in the 72-73c range. Still very acceptable, but yea...

Thanks for the heads up
 
ASUS sensor seems to to show/display somewhat lower than CoreTemp or RealTemp... still that seems like decent temps for the voltage and o.c. your pushing on air... maybe you've got a higher VID/low leak sample? .

Here is mine... Prime large FFT ambient temp ~22.3C:
 
PolRoger90 said:
ASUS sensor seems to to show/display somewhat lower than CoreTemp or RealTemp... still that seems like decent temps for the voltage and o.c. your pushing on air... maybe you've got a higher VID/low leak sample? .

Here is mine... Prime large FFT ambient temp ~22.3C:
Click to expand...

For some reason every time I use realtemp, it reverts my Multi to stock settings.

Dunno whats up with that.

Those are nice temps for sure

Man that is a badass voltage to hit 5ghz.

I r jealous!
 
DASHlT said:
For some reason every time I use realtemp, it reverts my Multi to stock settings.
Click to expand...

Are you sure this is an issue?

Keep in mind that overclocking on the 2500K isn't done by changing the multiplier on the base clock, it's done by changing the multiplier on the turbo-boost frequency. That means regardless of what your overclock is, it's normal and intended behavior for the chip to return to stock speeds or even lower when idle for power saving reasons. You would only have a problem if it's not kicking back up to full speed under load.

As far as your temperatures, you have neglected to inform us of your ambient temperatures, probably the biggest variable when it comes to temperatures. There is a big difference between temperatures with an 80F ambient and temperatures with a 50F ambient. Your location shows you as being in Spokane, Washington and it's currently in the middle of winter. Washington just experienced a huge snow storm, didn't it? I'd take a guess and say your ambient temps are a bit on the low side, which of course is going to make your CPU temps look amazing.

It appears that you're monitoring one temperature value for your entire CPU? Why wouldn't you measure individual core temps? Are you sure that you are monitoring the TJunction temperature and not the TCase temperature with that program? Monitoring TCase temperature will report a significantly lower value than monitoring individual TJunction (core) temperatures.
 
DASHlT said:
For some reason every time I use realtemp, it reverts my Multi to stock settings.

Dunno whats up with that.

Those are nice temps for sure

Man that is a badass voltage to hit 5ghz.

I r jealous!
Click to expand...

Have you tried CoreTemp?

Thanks! I just recently switched my bench table over to water (still a WIP)... Phobya Extreme 400 with Swiftech Apogee HD block:
 
^^^^ I agree, use coretemp or realtemp. My asus reports around 10c lower than those two programs. I was about to get angry with my water again, but then saw you using asus to monitor temps....
 
My old 2500K @ 5.0GHz @ 1.4V maxed around 70° C on an Kuhler 920 so 65° is not out of the question...

BUT YOU ARE RUNNING 1.5V YOU MADMAN! I have no idea why your system is so cool. Something must be wrong or you may have a golden chip. Why so much voltage? Won't work @ 1.4V?
 
Yes, if you are using AI suite you are showing the wrong temperature. AI Suite shows the hottest core temp minus an offset.
 
if anyone bothered to read forum posts. You would know that realtemp was bugging out on me which is why I was using the Asus suite one.

Now I was using coretemp, which still only showed 3-4c more then asus suite on my cores.
 
After doing some overclocking tonight. I noticed that I basically need 1.5 volts to hit 5ghz, compared to 1.4v for 4.8ghz. I do not think it's worth .1v for 200mhz.

Ambient is around 70-71f in the office.


48ghz.jpg
 
Last edited:
DASHlT are you running W7 SP1 x64 and what is your maximum sustained power reading in coretemp?
 
You guys see what the problem is right? Look at the ibt gflops....they are way two low. Sir to find out how good your chip actually is first install win 7 sp1 and make sure avx is installed. Run ibt afterwards, now your gflops should be around 130-140 and temps much higher. :). Also your power reported in core temp seems wayyyyy off. My 2500k @ 4.6 ghz hits 133w max at 1.35v with -.08v offset. It also gives around 125gflops on very high setting run (memory @ 1600). Gl sir
 
haris525 said:
You guys see what the problem is right? Look at the ibt gflops....they are way two low. Sir to find out how good your chip actually is first install win 7 sp1 and make sure avx is installed. Run ibt afterwards, now your gflops should be around 130-140 and temps much higher. :). Also your power reported in core temp seems wayyyyy off. My 2500k @ 4.6 ghz hits 133w max at 1.35v with -.08v offset. It also gives around 125gflops on very high setting run (memory @ 1600). Gl sir
Click to expand...

I have memory at 1333 and 2500k@4.6Ghz I'm only getting 111GFlops?!
This is on 2048MB, maybe I'll try 6500MB (1500MB is used by windows)

Could 1600 memory make a 14Gflops difference?!

I'm on +180mv (+0.18v) which gives me 1.4v in Linpack with added AVX, it does peak at 1.42v, but most of the time it's 1.4v.
 
grandpatzer said:
I have memory at 1333 and 2500k@4.6Ghz I'm only getting 111GFlops?!
This is on 2048MB, maybe I'll try 6500MB (1500MB is used by windows)

Could 1600 memory make a 14Gflops difference?!

I'm on +180mv (+0.18v) which gives me 1.4v in Linpack with added AVX, it does peak at 1.42v, but most of the time it's 1.4v.
Click to expand...

Hi, yup that sounds about right. When I run my IBT on standard settings (2000mb I believe), i get around 112-114 gflops. Faster memory and running the test longer will increse your gflops. :), Testing your gflops is always a good way to see if your processor is stable, if your gflops fluctuate a lot thats a sign of instability.
 
You must log in or register to reply here.
Back
Top