is my e8400 core temp too high or normal?

[mooglegoogle]

Here is a link to my 11hr orthos result: http://nimblemoogle.googlepage....JPG/oc-full;init:.JPG

Setup:
P5E with bios 0607
hsf - nirvana nv120 with as5
vcore in bios set to 1.175
vcore is 1.13 to 1.144 idle reported by cpuz and noted to idle at 1.13 in bios

My concern is the temps of core 0 and core 1 with cpuid monitor. core 0 is at 45C on boot and core 1 is 42C on boot and cputin is 32-33C idle. Not sure if it is normal for 1 core to be hotter than the other but they both end up capping at 56C under stress which is also odd how the higher temp core rises in temp less then cooler core.

Comparing with anandtech's heatsink review, http://www.anandtech.com/casec...howdoc.aspx?i=3268&p=7
I noticed the zerotherm Nirvana nv120 idle is pretty low at 30-34C in the 3.3 to 3.7ghz range, along with many other members posting an idle of 30-34 at a 1.2 - 1.4. If I am at 1.14 vcore, shouldn't my temps be lower than people with with 1.2+ vcore at the same speed of 3.6ghz?

I originally put 3 lines of as5 down the center, but thinking it was too much and was the cause of the high temp, I cleaned it off and reapplied 1 line through the center yet my temps remained the same after a few hours of testing. Based on the info from the pic, is the temp normal for a vcore of 1.14 or is there a way to reduce the idle temp?

Second, I was originally aiming for a 4ghz oc but would I really notice any real world performance by gaining another 400mhz to the cpu only? considering that I am already running at 450mhz mem and 1800fsb.

 

[AbnerDoon]

I wouldn't think it would be the AS5 it sounds like even with 3 lines you put very little on like recommended. I paste it on myself and have the same temps running @3.6ghz under Orthos with a Tuniq Tower 120 heat sink.

 

[zfooz]

I tried the same thing with my E8400.. One of my core temps is like 52 and other is 62 under maximum load/heat in prime95.
Mine's at 3.9ghz 455x 8.5 with 1.3v.

36 hours of orthos without an error so I'm happy.
I tried running it at 500x8 and it wasn't stable, and it wasn't really a whole lot faster. A little bit but nothing noticable. The RAM speed on my board is pretty slow (ABIT IX38) for some reason. Ram is running at umm.. something like 1090mhz.. It's PC8500 corsair ram but still latency is higher and synthetic benchmarks put it around same as a stock ddr3 machine.
I also took my heat sink off (Tuniq 120) and put on AS5 in different configurations and the temps are exactly the same. I think lapping it is the only answer to the temp differentiation , unless there's a problem with the cpu diodes themselves which I wouldn't be suprised.

Stable is better than speed IMHO.. If it can't run 24h of ORTHOS or more on blend tests it's not worth doing.

 

[OCGuy]

You cant look at other people's CPU clock speed / voltages and try to match temps. They are affected by too many variables...ambient temp, the airflow in your case, etc.

 

[mooglegoogle]

Noticed alot of people saying they idle in the 30s yet hit the 50's under which seems odd since I only hit 55C if i look at the values of each core seperaly and each core idles between 43-45C.

Using cpuid hardware monitor and going by the temp for cputin being 32C, gonna assume this is the same as the bios temp which reads the cpu temp as 32C when I idle there. while under load my cputin temp is 44C. Testing at 3.8 with a vcore of 1.18 increased max temps 2C and gave an error after 30mins. I could increase it some more and actually use 1.25 vcore and probably make it stable but don't feel its worth the heat so gonna use 3.6 at 480 1:1 and 1920fsb instead.