E8400 TEMP ISSUE,VERY DETAILED,looking for answers！！！！

[Coreyhm1]

Is it normal for my "cores" to be reading a 7*c difference. On load "Core1" jumps to 56* instantly while "core0" slowly increases in temperature

 

[imposter]

Pack data of 1/9/08 here... does that make mine "new"?

I am sorry i don't understand you

coretemp readings are off. why are you rmaing them?

 

[harlam357]

I am sorry i don't understand you

coretemp readings are off. why are you rmaing them?

I am sorry... I don't understand you. You said...

New e8400's will not work with coretemp.

then thehacker010 said...

Just so you guys know my pack date was 12/20/07, which would make it an "old" one.

then I said...

Pack data of 1/9/08 here... does that make mine "new"?

So the question here is... How are you defining "new"? Any and all E8400s? I think that's what you mean but tis not clear.

Basically it read like you were implying that an older batch is ok and a newer batch is not... but now I realize that is not what you were saying.

...and for the record, no RMA here.

 

[blazz]

E8400 with ASUS P5B Deluxe

I too was under the impression that the temps of the E8400 were going to be lower. I ebayed my X6800 and purchased a E8400 a week ago. I had my X6800 clocked at 3.61 using the unlocked top multiplier. I have my E8400 clocked at 3.61. I am using a Synth Ninja Cooler with a SilenX 120 fan. The X6800 at 3.61 & 1.425v ran at 41 idle and this E8400 @ 3.61 & 1.25 is running at 49-50 (Readings from Core Temp .95.4) I thought my temps would be lower with this proc. than the X6800. (Also with my Bios and this E8400, and my MB it will not boot with 4GB of Ram, get the BSOD.) Not sure whats going on there. Anyone else have that happen?

 

[Omsion]

I too was under the impression that the temps of the E8400 were going to be lower. I ebayed my X6800 and purchased a E8400 a week ago. I had my X6800 clocked at 3.61 using the unlocked top multiplier. I have my E8400 clocked at 3.61. I am using a Synth Ninja Cooler with a SilenX 120 fan. The X6800 at 3.61 & 1.425v ran at 41 idle and this E8400 @ 3.61 & 1.25 is running at 49-50 (Readings from Core Temp .95.4) I thought my temps would be lower with this proc. than the X6800. (Also with my Bios and this E8400, and my MB it will not boot with 4GB of Ram, get the BSOD.) Not sure whats going on there. Anyone else have that happen?

Your CoreTemp is reporting a 100 TjMax, while evreyone else has it at 105, which we think is correct (this also means that your temps in v0.96.1 would be 5° higher). Maybe we can go back to blaming CoreTemp

and all around

 

[satandole666]

I said it once I'll say it again...

I gave up on CoreTemp for these things. My temps are "stuck" at 40/38 until I get loaded with a moderate OC. I've seen CoreTemp hit 80c+ (92c once I believe) and didn't experience throttling.

My mobo reports an idle temperature of around 32-34c. When I boot in Windows, Everest reports a similar 32c. Not only do these values correspond with each other, they both scale with voltage and Everest scales with load. I've seen Everest report 70c and I got a little nervous. I have had zero reproducibility or predictability concerning CoreTemp's readings. Everest/mobo temps are the values I trust right now.

I remember overclocking before we had CoreTemp to tell us these more "accurate" temps. I have no problem going back to the old way of doing things, just have to realize that the mobo/everest temps aren't as lenient. 70c in Everest is bad, 70c in CoreTemp is just a walk in the park.

 

[forceman]

Well, I ran a little experiment the other night to check out my temps. My E8400 idles at 50+ no matter what voltage or speed it is at (tried stock, 3.6, 4.0) and loads up in the low 70s (coretemp temps - CPU temps in HWmonitor and speedfan are about 10C lower). I find it hard to believe that it idles at 50+ at stock voltage and speed so I decided to find out what happens at high temps. I pulled the plug on the fan of my Ultra-120 and ran Orthos in the strss CPU option to heat up the core. The temps slowly climbed up to 104C (on coretemp - Tjunction 105C) and then froze there (which makes sense I guess since the DTS is reading delta to Tjunciton and probably isn't designed to read negative numbers). Watching on Throttle Watch I got some weird dips in core usage right around the time it hit 104C but then seemed to go back to 100% and stayed there. CPU-z started showing weird fluctuating VIDs around that time also - jumping from 1.125 to .0825 and back. I kept it running and coretemp started showing weird temps, -20C and then 127C, then -20C again, so I went ahead and plugged the fan back in. I checked the heatsink before I plugged it back in and it was HOT.

So what did that prove? I'm not sure. It seems to imply that coretemp is reading the delta to Tjunction correctly, because it reported accurate deltas right up to the point it reached what it though was Tjunction (which may not be correct - maybe it isn't really 105C). I still find it hard to believe that my chip idles at 50+C, but maybe the Tjunction is wrong on Coretemp? If it was really 85C that would make my idle temps mid-30s C, which seems a lot more realistic. Or maybe these chips just run really hot.

 

[g14novak]

I jumped straight from the 1st to the 4th page on this so bear with me but...

Has anyone tried lapping them or do the ketchup test, assuming that Intel is shipping them with terribad IHS's?

If so, could we see some pictures of the ketchup test?

 

[forceman]

I lapped mine, didn't seem to make any difference.

 

[CreasianDevaili]

These are my stock readings.. havent tried to Oc yet..

Using a 9700 LED for the HSF btw..

CPU: e8400
Batch# Q746A516
Pack Date: 01/29/08

Core Temp reads: 39c on both cores. Shifts for a second between 38c and 40c, but pretty much stays put on both cores at 39c idle.

Then I ran prime95. The temps jumped up pretty fast to about 46. Now they stay there, sometimes for a second hitting 47c. Then as I watch, about 2 hours in, they sometimes go down to 45c. Both cores are being used and at 50%.

This is my first multi-core processor, so I am not used to how they should read for temps. I am trying right now just to see if my 9700 LED is doing it's job.

Now yesterday, right after i got it together, I ran all this and I was 42c on idle and about 50c'ish about a hour and half into prime95. So the temps seem to be going down the more I let this cpu burn in.

The cores are staying pretty much equal with core temp. Sometimes they differ by 1c or 2c but its only for a few seconds running prime and then they are the same again.

Just thought I give my temps and batch/pack to see if there is a improvement in readings from earlier batches.

 

[jason4207]

These are my stock readings.. havent tried to Oc yet..

Using a 9700 LED for the HSF btw..

CPU: e8400
Batch# Q746A516
Pack Date: 01/29/08

Core Temp reads: 39c on both cores. Shifts for a second between 38c and 40c, but pretty much stays put on both cores at 39c idle.

Then I ran prime95. The temps jumped up pretty fast to about 46. Now they stay there, sometimes for a second hitting 47c. Then as I watch, about 2 hours in, they sometimes go down to 45c. Both cores are being used and at 50%.

This is my first multi-core processor, so I am not used to how they should read for temps. I am trying right now just to see if my 9700 LED is doing it's job.

Now yesterday, right after i got it together, I ran all this and I was 42c on idle and about 50c'ish about a hour and half into prime95. So the temps seem to be going down the more I let this cpu burn in.

The cores are staying pretty much equal with core temp. Sometimes they differ by 1c or 2c but its only for a few seconds running prime and then they are the same again.

Just thought I give my temps and batch/pack to see if there is a improvement in readings from earlier batches.

You need to get the multi-threaded version of P95. If your only at 50% then your not stressing correctly.

 

[g0dM@n]

Okay, so I just got my E8400 fired up. It's an SLAPL (I wrote everything down, but no idea what other #s are important)... check out the temp readings I get at stock voltages (D-Tek Fuzion waterblock, D5 pump, Bix2 Rad, Liquid Metal TIM):

P.S. How can I tell what my default vcore is... Core Temp says VID 1.050v, but bios reads it as 1.15v, and since Core Temp seems to be way off for my temps I don't trust it.

 

[jason4207]

That VID should be accurate. I've seen some at less than 1v.

 

[LilBuddy]

I'm curious about my Vcore also...

Bios Setting: 1.305
Bios Reading: 1.20
Core Temp: 1.0375
CPUID Hardware Monitor: 1.19
CPUID: 1.192
Abit EQ: 1.19

What reading is correct?

 

[Omsion]

I'm curious about my Vcore also...

Bios Setting: 1.305
Bios Reading: 1.20
Core Temp: 1.0375
CPUID Hardware Monitor: 1.19
CPUID: 1.192
Abit EQ: 1.19

What reading is correct?

CoreTemp's is labeled VID, so that's essentially default vcore for that particular chip.
Throw in vdroop is why your readings are not 1.305, and apparently there's some more going into Windows. So ~1.19. If you want to be sure, throw a voltmeter on it.

 

[RangerXLT8]

IT seems like some folks are misguided about processor temps. Core temps are not the temperature that you need to worry about, CPU temp, which is measured at the heatspreader is what you need to worry about. The cores can take temperatures up to 90C and not throttle or be damaged. I'm in no way saying to run temps that high but merely making a point.

In response to anna original post, there is nothing abnormal about seeing a core temp increase by 15-20c in 1 second once the CPU is loaded. Then these core temps slowly rise up a few degrees as the CPU zone heats up.

 

[HousERaT]

I'm still going with the idea that coretemp is not reading these chips properly. I had a benching session this weekend and my temps were way below 32°C but the temp on coretemp was locked at 32 degrees idle. I don't know if this is a bios, sensor, software bug or what but something ain't right. Does anyone know a program that reads these temps accurately in windows?

 

[Trap05]

I JUST received mine with a pack date of 1/29/08 SLAPL What can anyone tell me about this chip?

 

[g0dM@n]

I JUST received mine with a pack date of 1/29/08 SLAPL What can anyone tell me about this chip?

forgive me, but how do you check the date?

CoreTemp's is labeled VID, so that's essentially default vcore for that particular chip.
Throw in vdroop is why your readings are not 1.305, and apparently there's some more going into Windows. So ~1.19. If you want to be sure, throw a voltmeter on it.

Well my VID from core temp was reading 1.050v, CPU-Z read 1.168v and SmartGuardian read 1.16vcore... this was at DEFAULT voltage.... so last night I set my vcore manually to 1.05 in BIOS and now CPU-Z reads 1.08 vcore and so did Smart Guarding I think...

So basically, I guess VID is being read wrong from Core Temp and my default vcore may likely be 1.15v instead!!!?

 

[Trap05]

Bottom RH corner of the sticker on the box.