SandyBridge 2D madness - let's tweak for the best 1m, 32m, pifast and wprime!

[SoF]

Nice time chew! Looks like 2500K isn't all way lost against the bigger cache of 2600K.

Yesterday I was testing some cheapo crappy MSI P67-C43. There is / was a contest on a german site where you needed MSI board so I went out to get the cheapest piece I can get. The bios is pita and bios up to .185 will allow 1,35Vcore only, new one will give you "amazing" 1,5V - there is no vccsa, pll or io voltage setting
My RAM wouldn't run good, I corrupted the bios several times, yeah it was fun ^^ But managed the "aussie basis reference time" of 6min 39s

6min 39s 500ms

[chew*]

Sof that run was full tweaks and highly tuned PSC. Still think 2500K is slower at min 2 secs CPC.


Jody I assume your on phase still?

I'm at max 20C cpu full load.......6C quite often.

Dino your on H70 water normal ambients?

Cheapseats what temp are you running the cpu at or keeping it under?

[dinos22]

yes as you can see in the video, i had 40C idle lol

[chew*]

yes as you can see in the video, i had 40C idle lol

yah you see where i'm going with this right?

Like a cold chip performance issue.

Anyway no progress with 32m yet so ran some maxmem last night.

Top 20 is easy with sandy.

[dinos22]

yeah time worth testing on air heheh

i put a maxmem time on the bot too :p touch faster heheh

[chew*]

yeah time worth testing on air heheh

i put a maxmem time on the bot too :p touch faster heheh

you can do better cpu speed boosts it a bit.

[3oh6]

Jody I assume your on phase still?

I'm at max 20C cpu full load.......6C quite often.

Dino your on H70 water normal ambients?

Cheapseats what temp are you running the cpu at or keeping it under?

nope, still on water. no point in insulating the board if it is 4 seconds slower than the M4E. i wanted to test on water first to ensure it wasn't cold
that was causing the memory clocking problems i had on the M4E.

under load the core running gets to 40C max.

as for sub timings, i too did one at a time and tested each. most don't show any gains, but tightening any of them didn't cause my time to slow.
this is something bigger than a single sub timing.

i was using a POS SSD and was thinking this morning as i got up if i should try a SATA. gonna also start messing with other BIOS's. but Brian is
right...this is retarded and not worth my time to be honest.

any chance you have a different rev of board Dinos? after the UD9 bullshyt, i don't rule anything out.

****edit - solved***
there it is. thanks for the images of the BIOS Dinos

it is the "No Execute Memory Protect"...needs to be disabled. flick of a switch

[chew*]

Good find, missed that myself. Confirmed it works for me to, shaved 4 secs no maxmem or waza.

Alex the compare has to be apples to apples, same OS same ram same HD same everything.

[dinos22]

woot good work mate

i automatically disable that crap, would have been scratching my head like all of you had i not lol

i'll ask the guys to move that option somewhere in the main tuning bios area so its easier to spot

****edit - solved***
there it is. thanks for the images of the BIOS Dinos

it is the "No Execute Memory Protect"...needs to be disabled. flick of a switch

[chew*]

Dino was having a good laugh, look at these idiots with no effeciency

Ok really need to get on one of my PI OS's to push this any further in the 5 gig range, that or add some cold to ram.

[dinos22]

badass

click to hear your theme song chew*

[Perica_Barii]

Just get this board from Gigabyte office in Serbia, and start playing with it...

Setup:
MB : Gigabyte P67A UD7 ( thanks to Gigabyte office )
CPU : 2600K batch#L040B694
MEM : OCZ BLade 2000 cl 7 8 7
HDD : WD Raptor 36GB
GPU : Gigabyte GTX260 SOC
PSU : ANTEC 1200W

Vqpi :1.14 , Vmem 1.72 all set in bios



will try lower multiplier and higher BCLK later....

[aristidis]

And one better from me, my CPU can't go higher with air.