With software such as OCCT which keeps the CPU at 100% for a longer time, I unfortunately had to make a step backward. No way to keep the computer responding more than a few minutes at 4.4GHz around 1.4V. Which means I would probably have gone at much higher voltage to talk about stability.Overclocking part
has been updated on the second post:
Originally Posted by BaK2BaK 4.3GHz@1.415V
1h OCCT / 3DMark11 / Cinebench passed
T°CPU max: 80°C (ambient 21°C)
Kill a watt max: 100W
Dark Rock TF max: 1000rpm
60mm PSU fan: 1500rpm
T° PSU max: 46°C
My safe voltage was arbitrarily set at 1.4V, already went beyond it to get a 4.3GHZ stable computer. Will see if I could go further as temp is still 'decent' (-20°C under TJ max).
But for now I am going to run it like this for a while seeing how it goes.
Do you guys think I can give more juice to that chip without risk?
I don't find any info on that about Broadwell CPU, but I see guys setting their Vmax at 1.5V for Haswell CPU. So I could still have a safe margin, maybe... Will ask Intel and tell you what they suggest about that.
Actually the CPU looked like being stable with 4.3GHz@1.37V on OCCT, but failed at finishing a 3DMark11 test. Which makes me think the iGPU was the problem.
I would have liked to try deactivating the eDRAM as mentionned here: http://www.legitreviews.com/edram-overclocking-on-the-intel-core-i7-5775c-broadwell-cpu_167425
But Intel Extreme Tuning Utility doesn't seem to monitor this on my rig and I don't know how to check this value. Going to ask Intel about this also...
Happy with the mod of the PSU fan, however at that speed it wasn't able to get rid of the heat and the PSU was slightly getting hotter and hotter.
I might use a knob at the back of the case to manually adjust the 60mm fan speed, though having the CPU running at 100% load for an hour will be rare on real life activities.