unityofsaints - Core i7 2600K @ 5583MHz - 6min 19sec 15ms SuperPi - 32M

[zeropluszero]

Tite. Wouldnt try to do hypers at 7-6-6 or 6-5-5 instead of PSC? Can you kick in that 56x multi for 32M? my old one I think i had to boot 1c1t to get it to jump from 55 to 56 booting. which is less than ideal for 32m but probably still ok.

[unityofsaints]

8 minutes ago, zeropluszero said:

Tite. Wouldnt try to do hypers at 7-6-6 or 6-5-5 instead of PSC? Can you kick in that 56x multi for 32M? my old one I think i had to boot 1c1t to get it to jump from 55 to 56 booting. which is less than ideal for 32m but probably still ok.

Yeah all solid tips. Hypers will do better for sure and 56x multi will set but doesn't complete. Probably heat-related. It's a good chip so I don't want to kill it by going above 1.65V on chiller. Will get the LN2 treatment soon-ish. Btw I had a few runs on Win 7 on 6 min 10 pace hang on loop 24 so the O.S. is definitely wankered. That's why I'm not spending any more time on it for now.

[zeropluszero]

yeh all good. keen to see what you can do :)

[yosarianilives]

Hypers are weirdly bad on sandy, I think that the bioses don't have subtimings set correctly for them. I've even seen where psc just inexplicably scores higher than hypers despite hypers appearing to overall being set tighter. 

[TerraRaptor]

I remember hypers required high latency boundary to work - killing the performance.

[Leeghoofd]

Hypers are ideal on P67 chipset boards, Z77 is not optimal (at least comparing Maximus IV versus V, 4dimms rule in efficiency clock per clock

[yosarianilives]

On 9/13/2019 at 12:15 AM, TerraRaptor said:

I remember hypers required high latency boundary to work - killing the performance.

Indeed, this was my problem on mve