What I ended up doing with my 4770k was delid, lap the bottom of the ihs, and use painted on mx-4, I might switch to CLP later. Then I found the proper settings for my cpu using linpack 10.3 and linpack 11.0. I used the linpack 11.0 settings as default multiplier/voltage (40x 1.20v) and set...
There have been several articles around the net about the fact that Haswell engineering samples performed much better than mass production Haswell. Not that many people run Linpack 11 so that may be why this issue isn't as well known as it should be.
This also only affects Haswell but that should be obvious from my previous post since SB/IB don't support AVX2 and I mentioned it was on the 4770K. Linpack 10.3 temps are around 20C cooler vs Linpack 11 on the same Haswell CPU. I'm also using custom problem sizes to maximize memory usage...
Actually its not good enough at stock speeds. Not even with a good aftermarket cooler. When running Linpack 11 (AVX2) with a Noctua NH-U12P SE2 my 4770K throttles at stock settings, due to the automatic overvolting the chips do.
Yes, Linpack is an artificial benchmark but once apps start using...
With further testing it seems AVX2 needs more voltage but perhaps a uniform amount of voltage until very high overclocks, ie:
3500 - 1.200v (1.175v failed) - (forgot to write down temp)
4000 - 1.200v - 97C
4100 - 1.210v - 98C
I changed the tests to be 4 runs of each 5k-60k in 5k increments. So...
I wonder how many of these 4770K oc'd systems really are stable after having run linpack 11 on my system. If I increase the problem size/dimensions on my system, eg 60000/60000, it will quickly throttle even at stock settings with a Noctua NH-U12P SE2. I've seen examples of people running at...
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.