surely

It's usually around 72~74c when gaming. It only gets to the 80s when running 3dmark 11. Also it'll hit 90 with furmark but I'm not going to run that ever again I think As a side note, with the 'modified msi FD' bios from using-your-4gb-gtx-680m-its-safest-full-potential I see the same temps but got a score of 7824 and much nicer FPS: NVIDIA GeForce GTX 680M video card benchmark result - Intel Core i7-3820QM Processor,CLEVO P170EM score: P7824 3DMarks throttling or not, I think this is an indication something is strange with the OCeditionNew clevo vbios? Or at east, the combination of that vbios and my card revision maybe? I ran each test a couple times to make sure it wasn't freak occurance

Hi everyone, I grabbed the vbios from 680m_Clevo_OCeditionNew.zip (I'm running a p170em and the clevo gtx680m it comes with) and did some testing in 3dmark with a few different clocks and noticed something odd: With the modified vbios I kept getting a 3dmark 11 score of ~6700, regardless of which core/ram clocks I picked - I think this one was at ~860 core (what ever the default was for that vbios) and 2280 ram: NVIDIA GeForce GTX 680M video card benchmark result - Intel Core i7-3820QM Processor,CLEVO P170EM score: P6745 3DMarks Temperature on that one hit 81c - not the highest I've seen but the fps was noticebly struggling. I flashed back to the original vbios I backed up before flashing the OC one and tested again, using just the sw overclock oh 719 + 135 = 854mhz and memory of 2280 and got a score over 7000: NVIDIA GeForce GTX 680M video card benchmark result - Intel Core i7-3820QM Processor,CLEVO P170EM score: P7038 3DMarks Is something wrong with this? My original vbios should be identical to the one the OC vbios is based on - 80.04.29.00.01 from clevo If the only difference between the original and the OC is the unlocking of higher core clocks, shouldn't these scores be closer if I use similar clocks? edit: On the stock vbios temperature max was 79, I guess that's comparable to the difference in the core clocks and makes sense, but the fps was noticibly higher on at least one of the tests, and visually there was less tearing and jumping

(forgive me for being a little off-topic) How do you do that? I had a look at XTU and I'm not sure what to change or the consequences of it, a little afraid of moving sliders to arbitrary values without understanding what they do. Can you point me towards some info on using XTU? Thanks

Wow, furmark manages to make it sit at 90c still when throttling like that? I don't think I want to run furmark any more

Ok, thanks. I'll have a read through the Clevo subforum to see what results people have had with repasting and modding the cooling on these. Thanks again

Cool, thanks. This brings up another point then, should I be concerned that I'm hitting 90c with even the default clocks when running benchmarks? Maybe my heatsink needs to be repasted properly or something. Seems to be ok if I'm just playing games though. eg borderlands 2 with everything turned to max hits a max temp of 80ish but I'm not sure its really as taxing as other games

Hi everyone, I've just receives a clevo p170em from originpc with a gtx680m and was interested in testing it's limits (safely) so I installed 3dmark, furmark, and nvinspector to have a play around. I found a guide elsewhere mentioning some safe values for core and ram overclocks without overvolting or bios modding to go past the +135mhz core. I noticed though that no matter what core clocks I set, either default 719mhz or oc 854mhz, when I run a benchmark the temperature slowly climbs up to 90c and then stops there - spot on 90c, never going any higher. I'm wondering, is it hitting some kind of throttling to keep it from exceeding that? Does this mean that effectively my changes to the core clock are achieving absolutely nothing? Also, what is considered safe temperatures for an overclock? 90c sounds pretty high but what can these things actually do in practice without damage or severely limiting the life of them? Thanks in advance for any input