[updated with temps]
My ICD7 just turned up today.
FYI I probably won't have time to test it tonight & I might not be able to do it until Saturday.(I'll setup the tables now).
PC spec:-
CPU - Q6600 @3GHz, vcore 1.304v (whilst loaded with DPAD).
HSF - Thermalright Ultima 90 with 120mm fan (I've already modded this setup by adding a washer under each screw head to increase clamping pressure, which is now moderate).
M/brd - MSI P6N platinium (Nvidia 650 chipset)
Case has 2 120mm fans.
Testing & setup notes:-
CPU loaded with 2 DPAD clients to keep the CPU as near 100% loaded as possible (1 doesn't quite) & then with Prime 95 v25.3a on small FFTs for max CPU temps. P95 temps taken 10 mins after starting it. Idle temps taken after a 10 minute cool down period.
CPU temps measured with Coretemp v0.96, case temp via speedfan & ambients with an old thermometer sat in my room.
The base of the HSF isn't flat & so doesn't make full contact with the CPU* so their is a larger than normal spread between CPU core temps (usually around 8C with DPAD & about 12C with P95 with small FFTs!).
*2 corners of the CPU don't touch the HSF (pic available if wanted) & so at some point soon I've got to lap it :disgust:. For this test however it'll be the same before & after, unlapped.
Before temperatures with Artic Silver 5 (installed over 2mths ago).
Ambient temp: 24.5C
Case temp: 28C
CPU temps C: _ idle _ DPAD ___ P95
Core 1 ______ 39/40 _ 54 _____ 65/66
Core 2 ______ 39/40 _ 54 _____ 65/66
Core 3 ______ 34 ____ 46/47 __ 56/57
Core 4 ______ 38 ____ 50/51 __ 59
After temperatures with ICD7 (about 1.5hrs after installation).
Ambient temp: 25C*
Case temp: 28/29C
CPU temps C: _ idle _ DPAD _ P95
Core 1 ______ 37/38 _ 50 ___ 61
Core 2 ______ 38 ____ 51 ___ 61
Core 3 ______ 33 ____ 46 ___ 56
Core 4 ______ 37 ____ 50 ___ 58
*By the time I took the P95 temps, ambient temp had raised to 26C & case to 30C, grrr damn central heating kicking in!
Wow!, despite that 1-2C dis-advantage for the ICD7 the 2 hottest cores dropped by a large 3-5C under load!:Q, odd that cores 3 & 4 dropped by only about 1C though......
Good to see the temp spread shrink anyway , maybe I can o/c more now?
I wonder if a large chunk of that drop is due to an improved application method?.....hmmm only 1 way to find out, I'm going to use the same app method with my AS5 (hopefully I'll have enough) & report back temps again, then we'll know.
After temperatures with ICD7 (about 5 days after installation).
Ambient temp: 23C
Case temp: 28/29C
CPU temps C: _ idle _ DPAD _ P95
Core 1 _______ 35 ___ 49 ___ 60
Core 2 _______ 36 ___ 50 ___ 60
Core 3 _______ 31/32 _ 45 ___ 54
Core 4 _______ 35/36 _ 49 ___ 57
Temperatures with AS5 using ICD7 application method (installed 12-4-08 14.50, about 2hrs after installation).
Ambient temp: 23.5C
Case temp: 27/28C
CPU temps C: _ idle _ DPAD _ P95
Core 1 ______ 35/36 _ 51 ___ 62
Core 2 ______ 36/37 _ 51 ___ 62
Core 3 _______ 31 ___ 45 ___ 54/55
Core 4 _______ 35 ___ 49/50 _ 57
Well a mixed bag here, firstly using more of the AS5 did help reduce temps compared to the original installation, the temp spread between cores dropped (8-9C down to 6C) but not quite as much as with ICD7 (mostly 5C).
Bearing in mind that the case temp was 1C lower than the ICD7 test, again cores 1&2 benefited more (by 1-2C,+1C) from ICD7 than did 3&4 (almost nothing, +1C) , I guess this has to do with the base of my Ultima 90 not being flat & ICD7 seems to overcome this better.
Ref case temp differences, it's currently 28C (ambient 23.5C) & running DPAD I'm getting 48, 49, 44 & 48C. So the extra 1C cooler I estimated ICD7 would of been, seems accurate.
I may do my GPU at a latter point if its not too much of a PITA to get the HSF off!
LIOUSS
You didn't note case or ambient temps.