e8600 not OCCT stable: too much or too little voltage?

[warbean]

I've been really happy with my new e8600 on a DFI Bloodiron. I thought it was stable because it passed 8 hours prime, 4.1ghz at 1.32v so I left it as my 24/7.

On a lark I decided to try OCCT and left it before I went to school today. I came back 10 hours later and my computer has shut down.

Does this usually mean a heating issue? Should I try less voltage then, to keep it stable? I realize 1.32 is somewhat high for a "modest" overclock like 4.1.

My temperatures I have been noticing are around 40c idle, 70c load, which appear to be in line for this model.

Running this on air on a xigmatek 1283.

Any thoughts appreciated....maybe even try increasing the voltage? It just kind of sucksthat I want 8 hours occt stable to be my golden standard, which means I can usually only run a test a day and not be able to use the computer to boot...

 

[OCGuy]

70C load seems kind of high for 4.1/1.32. I never see over the 50's @ 4.3 at near the same voltage.

 

[Saga]

I'm getting 65 load MAX running 4.2ghz @ 1.15v on a eVGA 790i using the stock heatsink. I was able to run 4.5ghz @ 1.24v but the temps shot up higher than I was comfortable with on load (85ish) which simply means the stock heatsink didn't cut it at that point.

Your temps are actually pretty low for that voltage, if I run 1.3v I end up with 90 degrees load (DANGER WILL ROBINSON) and really don't intend to push that kind of voltage again until I change out the heatsink. Do you really have to run the voltage so high, though? I'd bump it down to 1.2 at that OC and you will probably be more stable.

At Ocguy31, are you talking 50's load at 4.3? I was getting 65 degrees at 4.2! You must have swapped out the crappy stock heatsink, what did you use?

 

[warbean]

I will try slowly scaling back the voltage. It probably is not necessary.

I just feel a little disappointed that I paid for a premium cooler like the xigmatek s1283. Wasnt the point of doing that so I can run high voltages without worrying overheating? I just ran occt again and it was stable for 1.5 hours at 1.32vcore (bios) / 1.291 (cpu-z) so probably still have moreroom to scale back, though it did go as high as 76c (Realtemp)

I'm just not understanding how OcGuy can get more than 20c cooler than me at a higher clock and same voltage. That suggests a sensor is wrong or that I have my heatsink installed improperly.

What are everyone else's idle / loads with this chip?

 

[error8]

Originally posted by: warbean
I've been really happy with my new e8600 on a DFI Bloodiron. I thought it was stable because it passed 8 hours prime, 4.1ghz at 1.32v so I left it as my 24/7.

On a lark I decided to try OCCT and left it before I went to school today. I came back 10 hours later and my computer has shut down.

Does this usually mean a heating issue? Should I try less voltage then, to keep it stable? I realize 1.32 is somewhat high for a "modest" overclock like 4.1.

My temperatures I have been noticing are around 40c idle, 70c load, which appear to be in line for this model.

Running this on air on a xigmatek 1283.

Any thoughts appreciated....maybe even try increasing the voltage? It just kind of sucksthat I want 8 hours occt stable to be my golden standard, which means I can usually only run a test a day and not be able to use the computer to boot...

I have 70 C with my E7200 at 1,45 V in bios at 4 ghz with Xigmatek Red Scorpion which uses a low rpm fan ( max 1500 rpm ) opposed to the classic model which spins at 2200 rpm. So it's something wrong with your airflow, heatsink install or with the room temperature.

 

[warbean]

http://www.xbitlabs.com/misc/p...-e8600/temp-e0.png&1=1


This graph from an xbit labs review shows that the idle temeprature is around 39c going up to 70c at load, which is comparable to my own experience of around 401c idle, 75c at load.

So I don't think something is crazy wrong, though I guess I'm ujust reluctant to open the case and reseat the heatsink and fan if I don't have to. I'm pretty sure I packed that thing in pretty tight the first time.

So no one knows if this is aproblem where I needed to ADD more voltage to the chip to make it stable? (remmeber the whole computer turned off by the time I got home, so i really dontknow what the error was)

W

 

[AmberClad]

Did you check the logs that OCCT created yet? Do any of the charts show that a calculation error occurred?

 

[Miramonti]

Originally posted by: warbean
I've been really happy with my new e8600 on a DFI Bloodiron. I thought it was stable because it passed 8 hours prime, 4.1ghz at 1.32v so I left it as my 24/7.

On a lark I decided to try OCCT and left it before I went to school today. I came back 10 hours later and my computer has shut down.

Does this usually mean a heating issue? Should I try less voltage then, to keep it stable? I realize 1.32 is somewhat high for a "modest" overclock like 4.1.

My temperatures I have been noticing are around 40c idle, 70c load, which appear to be in line for this model.

Running this on air on a xigmatek 1283.

Any thoughts appreciated....maybe even try increasing the voltage? It just kind of sucksthat I want 8 hours occt stable to be my golden standard, which means I can usually only run a test a day and not be able to use the computer to boot...

My advice would be to rma it if still possible.

That said, I have one that reads temps even higher than yours does, so no its not the norm, but its not the first. I even lapped it, but all that was accomplished was a few degrees lower and a voided warranty. Idle at 1.16v was 48c.

It hasn't stopped me from cranking it tho. Currently its at 4.4ghz @1.42v, and can display temps has high as 89c under longterm full load (and with fans fully cranked). C'est la vie...if it dies I'll by a new one, lol, but I need the performance out of it.

Previous to the e8600, I had an e6850 (3ghz running @ 3.75ghz) and had 30c idle and 48c full load. My case has 4x120mm fans plus one 120mm fan on the scythe ninja 1100 cpu cooler. In fact, one 120mm case fan blows directly into the cpu cooling fan and the case is like a wind tunnel cranked, so I'm confident its a bad cpu in some way or another (but may just have bad temp readings, because I've never had a cpu run stable this hot.)

 

[error8]

Originally posted by: jjsole
C'est la vie...if it dies I'll by a new one, lol, but I need the performance out of it.

That's the spirit. :laugh:

 

[warbean]

Reseatedthe heat sink and fan. I can only conclude I have a hot chip. Idles at 40c and loads Prime at 95. OCCT Takes it as high as 73.

Yikes...