i7 920 + H50.....

[drstk]

i just got the corsair h50 and I'm testing it on 4GHz.

43-47c idle
im getting 83c on max load running p95 for about 30mins.

http://farm3.static.flickr.com/2516/4409154789_8e118484b1_b.jpg

i dont know exact ambient temps at this moment (hawaii)

HT is ON
191x21 multiplier
vcore @ 1.25V
PLL V @ 1.88V
QPI/DRAM V @ 1.21875V
IOH V @ 1.2V
ICH V @ 1.3V

should i stick with 3.6GHz w/ HT vcore 1.18125V (36-40c idle, 68c load) or 4GHz?

 

[shansoft]

I say that temp is fine enough..

I am using 200*20 and that require 1.32 for HT to be on, without HT would be a lot lower voltage and cooler..

why don't you try 200*20?

on my MOBO if I use 21 multiplier, it will decrease back to 20 whenever its on 100% full load, but not on anything lower....
does that happen to you too?

 

[Elmy]

I know you using that H50 might seem to be " COOL " because now im " watercooled " but in truth many air cpu coolers can do a better job than that thing.

 

[yang88she]

I know you using that H50 might seem to be " COOL " because now im " watercooled " but in truth many air cpu coolers can do a better job than that thing.

um...what? the H50 is a great CPU cooler and runs on par if not better than many air cooled solutions...and quieter

yes I did a personal comparison w/ my TRUE push/pull and am much happier w/ my H50

 

[drstk]

I say that temp is fine enough..

I am using 200*20 and that require 1.32 for HT to be on, without HT would be a lot lower voltage and cooler..

why don't you try 200*20?

on my MOBO if I use 21 multiplier, it will decrease back to 20 whenever its on 100% full load, but not on anything lower....
does that happen to you too?

aight. im doing 191x21 because i read that x19 and x21 is more stable then x20. but i didnt try x20, so i cant say it is. im going to run p95 over night to check for stability.

and when i run p95 it doesnt down clock. its steady at the 21x.

 

[drstk]

I know you using that H50 might seem to be " COOL " because now im " watercooled " but in truth many air cpu coolers can do a better job than that thing.

i didnt say anything about "h50s are better than air coolers"..... troll much

 

[p3sty]

um...what? the H50 is a great CPU cooler and runs on par if not better than many air cooled solutions...and quieter

yes I did a personal comparison w/ my TRUE push/pull and am much happier w/ my H50

I feel the same I had the same setup and I have been running 3-5c cooler with the H50

 

[BallerX]

Temps seem about right to me. My i5 750 @4ghz does about 75c max running linpac with the H50. I went down to 3.6 simply because I couldn't see a difference other than in canned bench marks and I can use a lower voltage.

Why do people think these things preform worse than air cooling? Its a pretty big copper block and a full 120mm radiator. The pumps flow quite well also. It had no problem besting my old Thermal Right heat sink in every way.

If you look around you find that these things are super easy to mod. Changing the rad to a dual and switching over to some more flexible tubing helps a lot. I like this a lot better than any air cooler I have ever used. The main reason being cleaning. I can just loosen up some screws and give a twist and it pops right off and I don't even have to move my case. Take it outside and blow the dust off. Most of the time I wash the radiator in the sink with a little dawn dish soap. We got smokers around here.

 

[Direfox]

I would love to see some perfomance numbers of stock H50 and one modded with a dual radiator. Anyone have any links?

 

[Zinn]

I've been out of the loop for awhile... so, the i7 series chips can safely tolerate 80+ celcius load temps? I thought these things automatically shut off when they hit 100 C. What is considered save temperature and voltage for this generation of chips?

 

[Potatocodone]

If the temps seems too high try turning the corsair h50 180 degrees.

I forgot what review site try this but their temps improve a lot.

 

[crazjayz]

I've been out of the loop for awhile... so, the i7 series chips can safely tolerate 80+ celcius load temps? I thought these things automatically shut off when they hit 100 C. What is considered save temperature and voltage for this generation of chips?

I'd say anything under 92*C is safe, which is when the CPU starts throttling down. 100*C is a safety precaution for the CPU, as extended use at 100*C will being to damage the chip.

 

[drstk]

If the temps seems too high try turning the corsair h50 180 degrees.

I forgot what review site try this but their temps improve a lot.

just turning it around or placing it on the top exhaust of the case? im running the h50 in a push/pull config btw.

i brought my i7 back to 3.6 and i realized my max temps are 10c-12c cooler than the cm v8. likeBallerX noted, running it at 4ghz i haven't notice that much of a performance boost unless I run benchmarks. So ill leave it at 3.6.

 

[FalconSS]

I have a 920 and an H50, I haven't Put in together yet due to lack of a mobo. So you guys are saying around 3.6-3.8 is the "sweet spot" currently for this proccessor?

 

[patric]

I have a 920 and an H50, I haven't Put in together yet due to lack of a mobo. So you guys are saying around 3.6-3.8 is the "sweet spot" currently for this proccessor?

A lot of people seem to be able to hit 3.6 without increasing the voltage. Both of my 920's are slightly undervolted and are 100% stable at 3.6GHz.

Also, rotating the H50 180 degrees did knock my temps down on my C0 about 5 degrees...it didn't do anything for my D0 build except get the hoses out of the way of my GPU.

 

[AMD T-type]

um...what? the H50 is a great CPU cooler and runs on par if not better than many air cooled solutions...and quieter

yes I did a personal comparison w/ my TRUE push/pull and am much happier w/ my H50

I'm getting better temps running my 920 @ 4.2GHz with the Cocage in the [H] heatsink roundup. 40C Idle / 73C Load

Ambient room temp of 23C



edit: I down-clocked to 4GHz and ran Prime95 for over 30 mins then let it idle for a few more

 

[hbidad]

Do you have your fan in a pull or push fashion? I originally had mine pushing and the swapped it around to pull to note a 5c decrease. Currently I get roughly 10C cooler with the same setup on both idle and full load. And I'm at 1.3v!

 

[AMD T-type]

Just using the stock fan in the push position. But I have a 120mm case exit fan not but a couple inches directly behind the heatsink so i figured a pulling fan would be pointless

I'm sure my temps are a bit lower since I'm in Chicago and he's in Hawaii, but its normal room temp in my condo

 

[hbidad]

AMD, my previous response was to the OP. If I recall, my temps was close to max at 75c polled from CoreTemp64. I'm @ 1.3.. well bios is set to 1.3, but CPU-Z tells me 1.288v (200x20)

Current idle temps are 32-40 across the cores.

What is odd is you have speedstep enabled, I don't. Your low temp @ 12x190 was 31c while mine is 32c at full speed.

I understand ambient effects this, but I'm not far off in Indiana and keep my office at 71F or better in the winter months. Although I do have a lame generic pull fan sandwiched. I'll need to test again with the this set to get a max. The last time I tested I primed for 20 minutes or so and held up the pull fan for about 5 minutes since I didnt have screws to hold it. The temps dropped a good 4-5c before my arm got tired. I since then took two screws from the other side to mount the other side.. /me to lazy to go to the hardware store.

 

[DeFex]

Just using the stock fan in the push position. But I have a 120mm case exit fan not but a couple inches directly behind the heatsink so i figured a pulling fan would be pointless

I'm sure my temps are a bit lower since I'm in Chicago and he's in Hawaii, but its normal room temp in my condo

if you can arrange it so the case fan IS the pullung fan then it does give you a little help. without any additional fan noise.

I have Casefan (still on case) held by 2 of the h50 screws to the h50, then the stock fan pushing on the other side.(held by the other 2 screws) although my 920 is an older stepping one and cant go anywhere near 4ghz, it runs nice and cool. (better then true, scythe mugen, and some other air coolers i tried)

 

[AMD T-type]

What is odd is you have speedstep enabled, I don't. Your low temp @ 12x190 was 31c while mine is 32c at full speed.

Why not? If you don't mind me asking.

Also, I'm not really concerned with my idle temps. I could care less if my idle temps were a few degrees cooler.

I may mess with it to see if it affects my load temps, but again, im rock solid at 4.2 while around 73C which is fine by me because i rarely do anything that pushes all cores @ 100%
I can play Battlefield BC2 for a few hours and the highest temps are mid 60's

 

[BallerX]

Just so you guys know. I started using linx because I found that Prime stable for 20 hours was a fail after about 5 passes of linx. Plus, it is easily 10-15c hotter. Linx is what I was using when I got an ultimate load test at 4ghz ,1.4v with a Max temp of 75c. You know your cookin when the radiator becomes warm to the touch. That's when I decided that 3.6@1.3v was for me.
Linx

 

[trudude]

My i7 idles around 38-40C and runs about 70-75C at max load. The H50 does a great job IMO.

 

[drstk]

Do you have your fan in a pull or push fashion? I originally had mine pushing and the swapped it around to pull to note a 5c decrease. Currently I get roughly 10C cooler with the same setup on both idle and full load. And I'm at 1.3v!

i have it set up as push/pull (push cool air in and pull) and exhaust from the top. so you flipped yours around and got better temps?

 

[HomelessDepot]

I've also got the 920 / H50 combo. Intake from the rear, over the rad.

~37 to 40 idle, ~75 max. Seems, like many water cooling setups, to maintain lower highs and higher lows .

3.8 ghz, solid as a rock.

-HD

 

[hbidad]

drstk, sorry.. was ass backwards. I originally had the single fan sucking air through the rad. I then switched it to push air through it.

 

[drstk]

ah ok. ic now