Which older air cooler should I use on 2600K?

Ice009

Weaksauce
Joined
Jun 22, 2008
Messages
93
Hey guys,

I'm about to upgrade an older socket 775 system (Q6600) that I have to a socket 1155 (2600K) based system, and while thinking about which Cooler to use I realized I still have a few laying around from socket 775 systems. As far as OC goes I'm only looking for a 4.5Ghz-4.6Ghz OC for a 24/7 rig.

The coolers I have are :

1. Thermalright Ultra Extreme 120
2. Noctua NH-U12P
3. CoolerMaster V8 (bought it second hand, but never got around to using this).

I don't have a socket 1155 mounting kit for any of these coolers, but I can buy one locally for any of these heatsinks, so just need to decide if it's worth using any of these coolers. What do you guys think? Is it worth saving some money and reusing one of these coolers?

I'm currently using the TRUE 120 on the Q6600 system and was thinking of buying the VenomousX BTK and reusing the TRUE on the 2600K system over the Noctua NH-U12P because the TRUE has 6 heatpipes Vs 4 for the Noctua, so I thought it would be the better choice. Not sure about the CoolerMaster V8 though as I never actually used it on a system before.

Also, what fan would you guys recommend on the cooler you choose? I have a spare GT 1850rpm fan that I can use.
 
Last edited:
At overclocked speeds the U12P and the TRUE will be practically identical I think. It really comes down to personal preference. The V8 on the other hand, is only a mid level heatsink, save that for lighter loads. But keep in mind, if you are not sensitive to noise all of these coolers should work, it's just that the U12 and the TRUE are more efficient.
 
I'd keep the True going on your Q6600... and buy a 1155 bottom bracket for the Noctua and use it with your spare GT 1850... a decent cost effective solution.
 
Thanks for the replies so far.

At overclocked speeds the U12P and the TRUE will be practically identical I think. It really comes down to personal preference. The V8 on the other hand, is only a mid level heatsink, save that for lighter loads. But keep in mind, if you are not sensitive to noise all of these coolers should work, it's just that the U12 and the TRUE are more efficient.

Do the amount of heat pipes usually have anything to do with performance? I didn't think the CoolerMaster V8 was a top end heatsink, but I find it disappointing that it has 8 heat pipes and doesn't out perform the other two, or at least be on the same level.

I'd keep the True going on your Q6600... and buy a 1155 bottom bracket for the Noctua and use it with your spare GT 1850... a decent cost effective solution.

I'm going to sell the Q6600, MB and RAM from that system and just put in a 2600K, Z68 MB and RAM while keeping the rest of the components in the system.

It's my internet box currently, but I also want to turn it into an encoding box. I have quite a few HD-DVDs that I want to backup to x264. I thought I should start backing them up since you can't buy HD-DVD drives anymore. That's the main reason for the changeover from the Q6600 to the 2600K. I also figured a 4.5Ghz OC would be what I should aim for if the main use of the rig would be for encoding.

So would you go with the Noctua or TRUE?
 
Thanks for the replies so far.



Do the amount of heat pipes usually have anything to do with performance? I didn't think the CoolerMaster V8 was a top end heatsink, but I find it disappointing that it has 8 heat pipes and doesn't out perform the other two, or at least be on the same level.



I'm going to sell the Q6600, MB and RAM from that system and just put in a 2600K, Z68 MB and RAM while keeping the rest of the components in the system.

It's my internet box currently, but I also want to turn it into an encoding box. I have quite a few HD-DVDs that I want to backup to x264. I thought I should start backing them up since you can't buy HD-DVD drives anymore. That's the main reason for the changeover from the Q6600 to the 2600K. I also figured a 4.5Ghz OC would be what I should aim for if the main use of the rig would be for encoding.

So would you go with the Noctua or TRUE?

Higher end air coolers usually have more heatpipes. Don't know much about the "V8" as I've never owned one. I do have a TRUE and a Noctua and have used both to cool 2600K's. They will do an adequate job cooling a 2600K up to average "daily type" overclocks of ~4.4/4.5/4.6Ghz. The True has closer fin spacing and two extra heatpipes and with a high speed/high cfm fans might be able to push o.c's (bench) a little higher but also with greater noise output. I like the mounting system of the Noctua. The close fin spacing of the TRUE can clog up with dust in dustier enviroments. I recently swapped a TRUE with push/pull fans for a TR HR-02 and a single 140mm fan in my wife's box as the TRUE seemd to require a little more frequent cleanings. Hopefully the HR-02's wider fin spacing will handle the dust better.
 
Higher end air coolers usually have more heatpipes. Don't know much about the "V8" as I've never owned one. I do have a TRUE and a Noctua and have used both to cool 2600K's. They will do an adequate job cooling a 2600K up to average "daily type" overclocks of ~4.4/4.5/4.6Ghz. The True has closer fin spacing and two extra heatpipes and with a high speed/high cfm fans might be able to push o.c's (bench) a little higher but also with greater noise output. I like the mounting system of the Noctua. The close fin spacing of the TRUE can clog up with dust in dustier enviroments. I recently swapped a TRUE with push/pull fans for a TR HR-02 and a single 140mm fan in my wife's box as the TRUE seemd to require a little more frequent cleanings. Hopefully the HR-02's wider fin spacing will handle the dust better.

Thanks for the info. Does the performance drop much when the TRUE gets a bit of dust between the fins?
 
Thanks for the info. Does the performance drop much when the TRUE gets a bit of dust between the fins?

Some...I noticed it... the load temps in RealTemp started to creep up... again this was on my wife's daily rig installed in a case. I like to run my own rigs on an open air test bench where it is much easier to see what is going on. Much will depend on how dusty your particular enviroment is and how often one uses a can of compressed air.
 
I've tested 3 coolers so far on my [email protected] ghz

H50- I was seeing load temps at just about 72ish
A70- Looking at 68ish
TRUE- 55ish

The TRUE is clearly a step above the other setups and a good match for the 2600k. I'm happy I bought the 1155 bracket. It was $23 shipped from frozencpu but well worth it. I was toying with the idea of getting the venomous X until I realized it really was just a TRUE with a shiny base and the new mounting bracket(which rules BTW).
 
I've tested 3 coolers so far on my [email protected] ghz

H50- I was seeing load temps at just about 72ish
A70- Looking at 68ish
TRUE- 55ish

The TRUE is clearly a step above the other setups and a good match for the 2600k. I'm happy I bought the 1155 bracket. It was $23 shipped from frozencpu but well worth it. I was toying with the idea of getting the venomous X until I realized it really was just a TRUE with a shiny base and the new mounting bracket(which rules BTW).

That's a good point about the new mounting bracket. The original socket 775 and 1366 mounting brackets suck, so it would not be hard to believe that the Venomous X mounting bracket alone makes a difference to performance. Also, good observation about the coolers looking similar. They've probably tweaked the VX to get a little bit better cooling performance over the TRUE, but maybe more of the performance comes from the mounting bracket itself.

What fan/s are you using on your TRUE?
 
Last edited:
I just got my 2700k build working about an hour ago. Used my TRUE Black with the 1156 mounting kit and I'm very happy with my results. I'm glad I didn't go buy another HSF.

I have a more typical OC as a starting off point, I plan on finding my voltage limits and such later, I just started off at 4.5ghz @ 1.34v


Idle temps are in the 26-29c range. Prime 100% temps are 56-62c.
 
Looks like I will use the TRUE. Thanks for the replies.

What fan/s are you guys using?

Edit : Should I just use the GT 1850rpm fan or how about the new Noctua NF-F12? I was looking at that fan as an alternative. Anyone used one? Gotta put the order in this week so if anyone can reply within the next day that would be great. Thanks.

Also the metal tabs on the bottom of the TRUE where the fan clips go on are broken. Anyone have the same problem? I've got a couple of those plastic fan holders that I bought a long time ago because of it, but not sure if you can use two of those plastic fan holders at the same time for push/pull?
 
Last edited:
So I went with the TRUE 120. Got the VX mounting kit and installed it with two GT 1850rpm fans. I was still able to mount the fans with the stock clips, you just have to have the fan's power cord at the top of the heatsink and you can still mount the fans by pulling the clips over a bit more than usual. Anyway, what I wanted to ask is, what is a realistic overclock with this setup and what do you guys use for stability testing?

I won't be doing anything really critical with the system, probably most stressful thing would be backing up HD-DVDs to x264 (not sure how stressful x264 encoding is compared to something like Prime 95?). I usually run 3-5 passes of LinX and then Prime 95 for about an hour and go from there. With my 1366 setup I ran LinX then Prime 95 for about 8 hours. Never had a socket 1155 CPU before so not sure how Prime 95 or LinX is as far as degradation on your CPU goes? Should I be more careful with this on Sandy Bridge? I read SB CPUs can fail quite often if put under too much stress with high voltage? What is safe volts for SB 24/7 setup?

Now onto the main issue, what temps should I be getting with my setup? and what is a realistic overclock for a 24/7 rig with this set up? When running LinX at 4.6Ghz my temps got to 90 degrees Celsius. I thought it might have been the mount so remounted and am getting basically the same temps with a 1 or 2 degrees difference on the remount. Also, one of the cores on both mounts when at 100% has a 10 degree difference. Is this normal with SB CPUs? On idle or mild loads the temps on the cores are pretty similar, but on full load Core 0 seems to be a lot cooler.

Edit : Ambient temp is about 28 degrees Celsius.

Edit 2 : Maybe I ask too many questions at once ;). Basically I just want to know are my temps too high for my setup? Does it sound like a mounting issue or would that be about the limits of the TRUE? I haven't tightened the pressure screw all the way as I read 70 lbs can damage the socket.
 
Using a v8 on my 2600k @ 4.8 ghz 24/7 with arctic silver 5 and temperatures are beautiful. Then again my hafx is a wind tunnel with a few extra modded fans on top of all stock fan spots filled. Oh and I changed the stock v8 fan. Idles at 27c and after hours of anything including crysis 2 ive never seen it get above 60. hardly ever gets even close to 60. The v8 came with a cpu mobo combo I bought from someone and I was gonna replace it but I figured i'd give it a go and I couldn't be happier with it. Looks badass too with a blue led fan in it.
 
So I went with the TRUE 120. Got the VX mounting kit and installed it with two GT 1850rpm fans. I was still able to mount the fans with the stock clips, you just have to have the fan's power cord at the top of the heatsink and you can still mount the fans by pulling the clips over a bit more than usual. Anyway, what I wanted to ask is, what is a realistic overclock with this setup and what do you guys use for stability testing?

I won't be doing anything really critical with the system, probably most stressful thing would be backing up HD-DVDs to x264 (not sure how stressful x264 encoding is compared to something like Prime 95?). I usually run 3-5 passes of LinX and then Prime 95 for about an hour and go from there. With my 1366 setup I ran LinX then Prime 95 for about 8 hours. Never had a socket 1155 CPU before so not sure how Prime 95 or LinX is as far as degradation on your CPU goes? Should I be more careful with this on Sandy Bridge? I read SB CPUs can fail quite often if put under too much stress with high voltage? What is safe volts for SB 24/7 setup?

Now onto the main issue, what temps should I be getting with my setup? and what is a realistic overclock for a 24/7 rig with this set up? When running LinX at 4.6Ghz my temps got to 90 degrees Celsius. I thought it might have been the mount so remounted and am getting basically the same temps with a 1 or 2 degrees difference on the remount. Also, one of the cores on both mounts when at 100% has a 10 degree difference. Is this normal with SB CPUs? On idle or mild loads the temps on the cores are pretty similar, but on full load Core 0 seems to be a lot cooler.

Edit : Ambient temp is about 28 degrees Celsius.

Edit 2 : Maybe I ask too many questions at once ;). Basically I just want to know are my temps too high for my setup? Does it sound like a mounting issue or would that be about the limits of the TRUE? I haven't tightened the pressure screw all the way as I read 70 lbs can damage the socket.

you're temps are pretty bad actually. What voltage are you running? IF it is set on AUTO you are probably pushing wayyyy too much voltage into the chip. My 2600k only needs like 1.28v to be stable at 4.4 and my loads are less than 60c running prime. If I leave the setting on auto in the bios it chooses something like 1.45 which is absurb.

The other thing I want to mention is that with the 1155 mount for your TRUE, you mention you haven't tightend it all the way due to pressure concerns. The screw actually works the other way, loosening the screw makes it apply more pressure. You might not have enough pressure to be effecient.
 
you're temps are pretty bad actually. What voltage are you running? IF it is set on AUTO you are probably pushing wayyyy too much voltage into the chip. My 2600k only needs like 1.28v to be stable at 4.4 and my loads are less than 60c running prime. If I leave the setting on auto in the bios it chooses something like 1.45 which is absurb.

The other thing I want to mention is that with the 1155 mount for your TRUE, you mention you haven't tightend it all the way due to pressure concerns. The screw actually works the other way, loosening the screw makes it apply more pressure. You might not have enough pressure to be effecient.

The voltage is set on Auto, but what kind of volts would manual require for 4.5 Ghz. Auto was about 1.32V on full load with LinX. I wouldn't think manual would get you much lower would it? I wanted to use the power saving features so that is why I have it set up like that. The rig is mainly an internet box that I will use to backup my HD-DVDs to x264 so I don't really want it running at full speed all the time. You can't put in a manual voltage and still have it down clock can you? My MB is an Asus P8Z68-V Pro.

Are you sure you have it right about the pressure screw on the TRUE? I am pretty sure you turn it clockwise to apply max pressure. It doesn't make sense doing it the other way. When you tighten the bolt clockwise it goes into the hole in the heatsink so that would apply more pressure wouldn't it as you tighten it all the way down? How would loosening it apply more pressure?

I really want to figure out why my temps are so high. Would an ambient temp of 28 degrees celsius be a reason?

Edit : What do you guys suggest? Should I tighten the pressure bolt more or try one of the other coolers. Not sure if I can use the Gentle Typhoons with the Noctua fan clips though.
 
Last edited:
Can anyone confirm if you can use any fan with the CM V8 and would Gentle Typhoons work with the Noctua fan clips of the NH-U12P?
 
try this for your voltage at 4.5GHz , set Vcore to offset mode and under that make it from -0.20 to -0.30 as every chip is different , it will downclock when not under load . forgot you have a 2600k , might need to turn off hyperthreading for these to work as i only have a 2500k (4 threads vs 8).
 
I've tested 3 coolers so far on my [email protected] ghz

H50- I was seeing load temps at just about 72ish
A70- Looking at 68ish
TRUE- 55ish

The TRUE is clearly a step above the other setups and a good match for the 2600k. I'm happy I bought the 1155 bracket. It was $23 shipped from frozencpu but well worth it. I was toying with the idea of getting the venomous X until I realized it really was just a TRUE with a shiny base and the new mounting bracket(which rules BTW).

Were you using the same fans? I found my H50 to perform as well as my true with my i7 860 when using the same fans. Literally, I saw no difference in temps.
 
Back
Top