i5 2500k overclock advice

[Veriitas]

Hi,

I just bought the i5 2500k and I want it to upgrade it to 4.5 Ghz.
I've been told that it is awesome :biggrin:

I just need a little bit of advice.

Will the "Cooler Master Hyper TX3" do as a cooler for my CPU?
Or do I need a fancy one? Any advice over upgrading 2500k to 4.5 ghz are welcome!!

I have never done it before.

Thanks in advance!

 

[stahlhart]

The subject has been covered exhaustively here in this subforum, and with a search you will be able to find discussion threads with suggestions.

That said, you should be fine with the TX3 -- if I remember correctly, this was the one Xbitlabs used to test overclocking for a review of the new (at the time) CPU line at their site, to make the point of how high it would go on air cooling.

 

[Veriitas]

Thanks for you reply!

I've read in a review that the TX3 makes an annoying sound when turned on high.
Do I need to put it real high to make it affective on a 4.5ghz overclock?

 

[stahlhart]

I don't really think so -- your limitations with a 2500K are probably going to be more about core voltage than temperature. If you've got good contact with the sink you should be okay.

You could switch to a quieter fan as well, if it ever got to that point, though your choices for a 92mm one will be limited compared to more widely used sizes like 120mm and 140mm.

 

[lehtv]

Coolermaster Hyper 212+ will take you to 4.5 in most cases but temperature will be over 70C IIRC. Dunno about TX3. If you want to stay closer to 60C you need to spend more money. E.g. Thermalright Venomous-X, Hr-02 Macho, Archon, Silver Arrow, almost anything from Noctua, Corsair H-series, Scythe Mugen 3, Scythe Mine 2. Or since you're in Europe, you could put your money on Alpenfohn Matterhorn. It looks fantastic IMO.

 

[Crap Daddy]

I have the TX3. You'll be fine.

 

[Veriitas]

Thanks all

 

[mrjoltcola]

FYI on the H60. I have it on my 2600K. At 4.5Ghz my cores get to 71-75 degrees C under load test. At 4.6Ghz it seems nearly identical. I run into stability issues otherwise regardless of temps, so while I'm glad I saved $$ with the H60 (vs the H80), if you want to run < 70 degrees, the H60 might not do it for you and you might consider the next bump up. It is very quiet and easy install though.

EDIT: I meant the Corsair H60

 

[Idontcare]

FYI on the H60. I have it on my 2600K. At 4.5Ghz my cores get to 71-75 degrees C under load test. At 4.6Ghz it seems nearly identical. I run into stability issues otherwise regardless of temps, so while I'm glad I saved $$ with the H60 (vs the H80), if you want to run < 70 degrees, the H60 might not do it for you and you might consider the next bump up. It is very quiet and easy install though.

EDIT: I meant the Corsair H60

Is that LinX? How many threads?

 

[mrjoltcola]

It was prime95, 64-bit.

I didn't try much harder, since my goal was only to break 4.0Ghz and run stable and cool long-term.

Is that temp typical for the H60 on the 2600K?

 

[Idontcare]

It was prime95, 64-bit.

I didn't try much harder, since my goal was only to break 4.0Ghz and run stable and cool long-term.

Is that temp typical for the H60 on the 2600K?

That's why I was asking, to see if it was comparable to what I am seeing with the H100 on 2600K at same clocks.

Two things to know about Prime95 and 2600K's...if you want to test your "stability at max temp" then you must set the thread count to 4 instead of accepting the default 8. 8 threads will result in lower temps than 4 threads.

Second thing is you must test explicitly with "small FFT", not blend and not large FFT, if determining your temps and stability are the goal.

Also, even with 4 threads and small FFT, Prime95 is not as exhaustive of a stress-tester as LinX (also w/4 threads). Just be aware of this.

That said, for prime95 I think your temps are as expected for the H60.

 

[jackstar7]

Also, on the H60, is that the single stock fan?

 

[mrjoltcola]

That's why I was asking, to see if it was comparable to what I am seeing with the H100 on 2600K at same clocks.

Two things to know about Prime95 and 2600K's...if you want to test your "stability at max temp" then you must set the thread count to 4 instead of accepting the default 8. 8 threads will result in lower temps than 4 threads.

Second thing is you must test explicitly with "small FFT", not blend and not large FFT, if determining your temps and stability are the goal.

Also, even with 4 threads and small FFT, Prime95 is not as exhaustive of a stress-tester as LinX (also w/4 threads). Just be aware of this.

That said, for prime95 I think your temps are as expected for the H60.

Maybe this weekend I will rerun my tests with your exact specs above, then I'll report back specifics.

 

[mrjoltcola]

Also, on the H60, is that the single stock fan?

Yes, the stock radiator fan that comes with the kit. Was a bit heavier than the Corsair fan that was in my case (I have a 600T Graphite). What about your fan?

My goal with liquid was:
1) Quiet
2) Simple install (funny, but the Corsair kits are simpler than some of the air kits to install).
3) Hoped to keep it 70 degrees

First #1 and #2 accomplished; #3, well, not at 4.5Ghz. I guess I am slightly disappointed, but not performance wise, just based on what I read. The temperature expectations were all in my head based on reading. Understand that I don't build a system often either, every 4-5 years maybe, so I can't even remember what good CPU temps are between builds. My old rig is a Core 2 Extreme X6800 Conroe from 2006 at stock frequency. But I'm having a bit of buyers remorse, wishing I had tried the H80. Maybe I could try a different radiator fan?

 

[aclim]

You could try a different fan and also try a push/pull configuration if you have not already done so. I would like to know what a nice Noctua fan would do in place of the supplied Corsair ones.

I have been in between an H60 and H80 for my new build. It's the last thing I need to buy. I would go with a H100, but don't want a dual radiator in my case. I will be using a Corsair 400r...