Sign in with
Sign up | Sign in
Your question

I7-2600k + P8P67 PRO + TurboV EVO OC = Yoiks!

Tags:
  • CPUs
  • Evo
  • Overclocking
Last response: in Overclocking
Share
January 24, 2011 7:31:23 PM

Hey guys.

First time builder. Got my system up and running and decided to let the Asus TurboV EVO overclocking have a go at it. According to their software it was perfectly happy running up around 5Ghz but... and here's where the question comes in... the CPU temps were getting up there so I killed the OC (immediately after seeing it getting hot). I'm wondering what the limits are? Is it normal that the system would be stable but that the heat would be the issue? And finally: did I damage anything (seems to be running fine – I set it back to defaults)?

Here are the details:

I let the TurboV EVO software do it's thing and it gradually worked its way up to 5Ghz but I noticed that the CPU temp was getting up around 80+C and the Vs around 1.4 so I killed the OC.

Hyper 212 plus
i7-2600k
Asus P8P67 PRO
4GB (2x2) G.Skill 1600 1.5V 9-9-9-24
Corsair 650TX
Antec 300 case

Thanks.
-Josh

More about : 2600k p8p67 pro turbov evo yoiks

a b à CPUs
a b K Overclocking
January 24, 2011 7:55:45 PM

No, you didn't damage anything. The chip would have auto-throttled itself had it hit the temp ceiling.

If you can, try adding a second fan to the 212+ cooler. Then it might handle higher CPU voltage like that. Otherwise, you'll have to manually overclock and keep the voltage lower so you don't approach the thermal limit.
m
0
l
January 24, 2011 8:40:44 PM

Will a second fan really make that much difference? I'm reading that it's maybe 2-3C cooler? At normal OC (4.4Ghz say) It's stable around 60-63C but 80+ just seems out of control. I guess the cooler isn't quite there... :??: 

On a related note: if I get a duplicate fan for the Hyper 212 Plus can I just get a splitter and plug both into the CPU fan port (for power and RPM)? Or, do I need to get one to pull power from the PSU and use the CPU fan settings solely to drive RPM?

m
0
l
Related resources
January 24, 2011 9:25:26 PM

Even though your built in software says 5Ghz shouldn't be a issue, I personally think that it probably is bit much. Unless you OC'd it manually yourself. Because honestly in my own personal experience with OC'ing systems I've noticed that most programs that are supposed to help you overclock usually end up using some pretty strange settings. Whenever I had just sat back and taken the time to slowly tweak everything bit by bit I'm able to get much higher than using a OC programs preset built in settings. I haven't finished my new i7 2600k build yet but once I do (which I'll be done in about 2 weeks) I'll let you know how high I was able to go with that specific CPU. Although from what I've read most people are able to get above 4.4 - 4.6 Ghz with almost no problems as long as they have some sort of aftermarket thermal solution/heat-sink etc... Even without any aftermarket stuff though people have been able to achieve 4.0 - 4.4Ghz without sacrificing any system stability...
m
0
l
a b à CPUs
a b K Overclocking
January 24, 2011 9:33:53 PM

Yeah, the difference would only be 2-4ºC in adding a second fan. Maybe the program was deliberately running the CPU that high to determine what the max OC was. You'd have to ask someone else who had that program though.

You can get a splitter, but you'll need to remove the speed-sensor (yellow) wire from one of the fans. The mainboard would get confused if it received two speed signals on one connector.
m
0
l
January 24, 2011 10:01:05 PM

ComputerNovice said:
Even though your built in software says 5Ghz shouldn't be a issue, I personally think that it probably is bit much. Unless you OC'd it manually yourself. Because honestly in my own personal experience with OC'ing systems I've noticed that most programs that are supposed to help you overclock usually end up using some pretty strange settings. Whenever I had just sat back and taken the time to slowly tweak everything bit by bit I'm able to get much higher than using a OC programs preset built in settings. I haven't finished my new i7 2600k build yet but once I do (which I'll be done in about 2 weeks) I'll let you know how high I was able to go with that specific CPU. Although from what I've read most people are able to get above 4.4 - 4.6 Ghz with almost no problems as long as they have some sort of aftermarket thermal solution/heat-sink etc... Even without any aftermarket stuff though people have been able to achieve 4.0 - 4.4Ghz without sacrificing any system stability...


I dunno. I think it was pretty surely doing 5+ as it was also showing that in other sensor apps but who knows. I got 4.4 using only the multiplier and no other changes. I noticed that the OC software upped the BLCK to 103 and the CPU voltage to around 1.4. Other than that, all seemed stock (including DRAM volts, etc). Then again, i didn't do a ton of snooping around - reset it back to stock pretty quickly.
m
0
l
January 24, 2011 10:04:45 PM

Leaps-from-Shadows said:
Yeah, the difference would only be 2-4ºC in adding a second fan. Maybe the program was deliberately running the CPU that high to determine what the max OC was. You'd have to ask someone else who had that program though.

You can get a splitter, but you'll need to remove the speed-sensor (yellow) wire from one of the fans. The mainboard would get confused if it received two speed signals on one connector.


Yeah it basically did a ramping up cycle: ran at say 4.4 and did some stress testing for 1 minute, upped it to 4.5, test and repeat until it got up to 5.0 (I was expecting it to crash way before that during stress testing). During the stress test is when I saw the heat really getting up there. I'm tempting to see if thermal grease is good but it has good temps otherwise: idle at 30 and stress (at 4.4 around 60). Asus tech support said max temps for the board are 65C... sound like they're covering their... err... asus. :na: 
m
0
l
January 25, 2011 12:21:38 AM

Yeah man honestly I don't know what else to tell you since I personally wasn't their myself, and since sadly my i7 2600k build won't be done for a couple more weeks since I still need to order a new case and new SSD for the system. I won't be able to offer any actual first hand knowledge. Because from everything I've read Sandy Bridge CPU's don't use the standard method (or the old method) of using BLCK overclocking to OC your CPU. Since Intel now put the clock generator on the CPU itself now (rather than the MOBO) it makes overclocking using the old method of simply raising the blck frequency no longer effective because now when you raise it you also raise the frequency of SATA and PCIE controllers as well as the SNB CPU itself.

(According to what I've read on anandtech, since I haven't messed with any of this stuff yet myself I can't offer any first hand knowledge)
m
0
l
February 24, 2011 1:57:33 AM

Getting 5ghz is easy on a 2600k system. All you need to do is adjust the multiplier, cracking 5ghz is the problem. The only way I have been able to do it is by setting the clock to 103, multiplier of 49 to 5.05.
m
0
l
February 24, 2011 2:37:30 AM

Yeah. It looked like the ASUS auto OCer would have taken it way up there. I think the big issue for me was cooling. How are you cooling? You doing this on air?
m
0
l
a b à CPUs
February 24, 2011 3:12:31 AM

I would say the big issue is actually voltage. The highest I have seen recommended is 1.375 V, and that is 1.4, so reasonably more.
m
0
l
March 1, 2011 2:13:41 AM

No, mine is cooled with a 120, I want to get a bigger one though.
The highest voltage i have seen on mine is 1.4 ish, it gets scary hot at 5GHZ, no real reason for it to be honest.
m
0
l
a b à CPUs
March 1, 2011 6:32:23 AM

No, the problem with high voltages is not the temperature, it is that the voltage will degrade the processor very quickly.
m
0
l
March 1, 2011 12:55:39 PM

You might wanna trying the OC with the side panel off. That way you can rule out if you have effectively made a convection oven with your case, if the temps are lower then look at improving airflow in your case as the ambient temperature inside your case should be about whatever the temps in your room are (without the PC on of course.) IE, dremel out all the metal mesh that covers your fan slots on the sides/front/back of your case. Also if you havent already done your cable management, might wanna get started on that as you want the optimum airflow as possible without ANY restrictions. as preferlinux stated the big issue is what your volts are running at, they will degrade your hardware faster/quicker more than the heat will as your chip/chipsets will (I should probably say SHOULD) throttle back or power down completely.

Also, are you running SSD's or SATA? SATA drives get pretty hot and if you have a front fan blowing inwards, it will blow across these drives(for cooling of the drives of course) and cause the heat to go into the case, which in turn increases temps inside your case. If you do, you may want to find the right configuration where this is negated as much as possible.

Good luck :) .
m
0
l
a b à CPUs
a c 100 K Overclocking
March 1, 2011 2:24:10 PM

I would first suggest considering why you're overclocking. Do you need the extra speed? Or is it more to see what that chip can do? If it's to push the chip to the limits, then OC it yourself manually in the BIOS. OC programs don't exactly care if they push too much voltage or temp, it'll just wait till there's a crash. Do it yourself, be safe, and you can probably hit a similar speed with lower voltage and temp.

Besides that if you're just OCing for a little extra performance which isn't really "needed" then just get it to a point where the temps and voltages are both decently below their max and be happy with that...
m
0
l
March 1, 2011 6:18:32 PM

PreferLinux said:
No, the problem with high voltages is not the temperature, it is that the voltage will degrade the processor very quickly.



This doesn't make sense. Of course one of the problems is heat! The temp on my 2600k can hit 70C when OC'd to 5GHZ, but barely cracks 50C at 4ghz. Do some research, any overclocked item increases in heat.

Maybe I misunderstood the previous statement, the Volts are the major problem, but the heat is a concerning factor too.
m
0
l
a b à CPUs
a c 100 K Overclocking
March 1, 2011 6:36:21 PM

brohde88 said:
This doesn't make sense. Of course one of the problems is heat! The temp on my 2600k can hit 70C when OC'd to 5GHZ, but barely cracks 50C at 4ghz. Do some research, any overclocked item increases in heat.

Maybe I misunderstood the previous statement, the Volts are the major problem, but the heat is a concerning factor too.


Well it's true that both cause issues, voltage is the worst for sure. For example you could try setting it at 2V, and before it has a change to really do anything the CPU will be toast. Have you seen those old OC benchmarks of i5s and i7s where they fried the bottom of the CPU? That was due to voltage. It'll basically "explode" the electronics. You can also check out electromigration on why it can degrade faster due to voltage.

I'm not really even sure why heat kills a CPU when most aspects of it should have pretty darn high melting points. I mean GPUs don't run into issues until over 100C. I suppose heat reduces electric conductivity so maybe it just compacts the issue... But anyway, while heat can cause issues I don't think it'll degrade the CPU too fast, especially since those max temps are only reached during stress testing and under normal operation the CPU is much much cooler. But the voltage won't change like that. It might fluctuate a little bit, for example gets reduced at idle or increased with turbo but under normal operation the voltage will always be roughly the same and that'll be causing wear all the time.
m
0
l
March 1, 2011 6:58:21 PM

This is true, I do see your point. My worry is for my 2600k that cranking it up to 5ghz, and hitting 75C at 1.4ghz is pretty intense lol
m
0
l
a b à CPUs
a c 100 K Overclocking
March 1, 2011 7:36:35 PM

Right, well first off why would you need to push 5ghz? I would say do it to run some benchmarks and prove how ridiculously fast the CPU is. But for daily? Maybe see what you can get out of it at 1.3Vcore or less? Clock for clock the SB CPUs are faster than last gens and nobody complained at a "mere" 4ghz lol.
m
0
l
March 1, 2011 10:09:04 PM

I am not going to argue about the 4ghz, it was purely for benchmarking
m
0
l
a b à CPUs
a c 100 K Overclocking
March 2, 2011 2:54:39 AM

Right, well fair enough. I would take what the TurboV gave you as a start. From there, go to the BIOS yourself and tweak it to try and bring the voltage and temps down. TurboV isn't exactly running full stability tests.

Unfortunately I haven't seen a P67 BIOS yet, my experience is with my own i5 750 on a P55 board. I don't know what voltages you have to adjust, but with mine Vcore and VTT/IMC/QPI (name depends on mobo brand) could be adjusted to affect the core speed. Raise one and you can drop the other. The key is finding the best balance.

Now maybe you don't have that. But I'd say try 5ghz at, maybe, 1.35V? If it doesn't work 1.36, then 1.37... etc. If it really requires 1.4V and the heat is too high it's up to you if you want to bench it or not. I've never taken my CPU right up to 1.4V. Mine is set at 1.36V Vcore for daily use, but that's with Load Line Calibration off so that means under load it's actually only 1.28V. Not bad. That's at 3.75ghz with turbo boost enabled (177x21 up to x24).

But, when I did my 4.13ghz OC I turned on LLC and took it up to around 1.36 or 1.37V... I still didn't hit the limit though because my temps were just over 80C under burn testing and I'm not comfortable going over that so for me, 4.13 is probably my max. Now, if I had water cooling and could keep my temps below 80C I would try 1.4V and also take my VTT up to 1.35V or so too...

So yeah, just saying, you should try lowering the voltage and testing and see what you can do. Those OC programs are ok but not as good as a human.
m
0
l
March 2, 2011 1:21:08 PM

Wolfram, you are correct. I wouldnt clock over 1.5v as it voids your warranty but, if you so choose to run at higher volts and temps and you are happy; then go for it. Just be prepared for consequences if/when it fails.

Heres the data sheet btw: http://download.intel.com/design/processor/datashts/324...

m
0
l
March 8, 2011 11:31:40 PM

Awesome info guys. Thanks! The autoOC kind of spooked me enough that I'm keeping it much more tame these days. Someone asked about motivation... I was just curious how far it could go and be stable. I guess that wasn't the limiting factor ;-)
m
0
l
!