Newbie: 920 @ 3.6ghz with EZ OC, worrying number..

[noktek]

Hi everyone, here probably a very boring and overasked question, but still...
i tought id overclock my 17 920 to the lower EZ Overclock setting my asrock x58 extreme allows me to, 3.6 ghz.
i never overclocked anything before, but since everyone says theres no risk and its really easy etc etc i tought id squeeze something more out of my cpu. I am not interested in extreme overclocking, i only want some more performance for my 3d renderings.

I have a Noctua 2 cooler for my cpu, so i guess on that side i should be okay, and 4-5 extra vents inside my case.

so i downloaded Speedfan to check the temperatures and i get a worrying little icon of a flame saying stuff well beyond 100 C°, in the picture here it says 127.. Also, i have no idea what this temp1 temp2 temp3 things are.. probably different spots on my grafic card getting real warm... or not?

can someone with a bit more experience than me comfort/warn me on what im doing?


Uploaded with ImageShack.us

 

[noktek]

also, my eight cores seem to reach temperatures around 68-72 degrees while rendering at full speed for 5 minutes. it seems stable. do you think if i keep rendering for longer the temperatures are going to rise or stay stable? is 70° acceptable?

 

[Rustyy117]

I don't think speedfan is reporting that temperature correctly, as that kind of temp would cause thermal shut down (PC would turn itself off in an attempt to save itself from breaking). Having said that, your GPU's VRM's (voltage regulator modules) can get upto around that sort of temp before causing thermal shut down. Unfortunately I don't know what temp 3 is, my guess would be a motherboard sensor (perhaps south bridge).
You only have two CPU core readings (Core 0 and Core 1) when you should have 4 readings as your core i7 920 has 4 physical cores and 4 logical cores (you won't get temp readings for the logical cores.) I would suggest you try hardware monitor or real temp to monitor your CPU's temps.

5 minutes of rendering won't tell you if your CPU is stable or not. If your doing work that has any value to you on your PC then I'd highly recommend you stress test your CPU's overclock for a minimum of 12 Hours, You can use programs like OCCT or Prime 95 to stress test. If you were to get a Crash/BSOD (Bluescreen of death) whilst rendering or doing anything really then you run the risk of corrupting the data you were working with.

70C is fine, I'd expect whilst stress testing for 12Hours it'll reach closer to 80C which is still acceptable, but at the higher end of my max temp recommendation, I certainly wouldn't allow it to go above 90C.

So you can loose data from unstable overclocks and if you let your CPU run at 90C+ for too long you'll kill your CPU and maybe even your motherboard, so whoever told you there was no risk is lying.

Hope that helps.

 

[noktek]

I tought temp3 was something GPU related, as the three temps are listed right after the GPU... anyway, that value seems to be a bit of a joker, as it goes up and down and is now been stable on 85 degrees for the last half an hour (in which i had not used my graphic card at all, as 3dsm seems not to be cuda able).

ive beenr rendering for half an hour now and the temps seem to be stable.. for the kind of work i do i rarely need such long rendertimes, so i guess im fine on that aspect.. . if i grow extra paranoid in the next couple of hours ill let a stresstest run over the night

here a link to my grafic card, http://www.msi.com/product/vga/N275GTX-Lightning.html#/?div=Overview

i tried to double proof the number of temp sensors it has but no infos about it.. i thought since its a bit of a "CrAzY" OC card with "CrAzY" components (at least they say so on the website) it might really have some superhigh temp components..
anyway, the official MSI utility confirms the 43 degrees of fanspeed

 

[Rustyy117]

Like I said before use Real temp or Hardware monitor, Speedfan doesn't look like it knows what its doing (127C on a sensor and 8 Core sensors on a 4 phyical core CPU)...

 

[noktek]

hardwarre monitor detects that weird temperature aswell... i really wonder what kind of component it might be....

 

[Rustyy117]

im pretty sure its a sensor on your motherboard, and I'm also pretty sure its faulty as you'd be smelling burning PCB at 127C

 

[Rustyy117]

I had my old 470 hit 100C. It was pretty sweet

lol, always fun times, I once got my HD5850's VRM's to 125C and my core i7 920's cores to 100C. Both were fine though.

 

[noktek]

opened the case and moved my hand around, didnt sense any kind of vulcanic warmth, so yeah, its probably faulty

 

[mkchampion14]

what the heck happened to the SSD? it was at 1C min. and 80C max???
everything else is fine, and the 127C is definitely a bad sensor, you can actually check by touching the heatsinks on your mobo and seeing which one gives you a third degree burn lol
also, i7's are made to survive 100C temps, though the highest long term temperatures recommended by Intel is 67.9C...

 

[noktek]

i dont know, usually the SSD has no temperature displayed at all