Leeghoofd

he's just a shy guy

another FM fail toooooooooo

Good luck Jerre, I sadly can't compete in this round as I don't have any of the older AMD stuff here only witht he 1090T I can try to reproduce a good run , though no idea how pcmark04 works

you can flame at Microsoft, game developpers, driver teams too then Chiller... you will have a busy day on the forums... heck a busy life... Nothing is perfect at release date, nor can it run 100% stable on all possible hardware/software combos from day one... that's eutopia !!

seems a bit up and down with the speed, sometimes I can't get access, sometimes front page and co works fine, forum doesn't... needs some time to settle but looks fab... totally refreshing !!

the one and only, otherwise no piccie like that

Blind... pretty cool he

Nice find Petri, awesome find... does it have Sayaas signature somewhere ? Maybe the best 790 SLI ever made

it's footy now

K404s scores on the first 3 tests scores far better then you, just in nature you pOwn him hard (with far lower clocks) Tweaking via OS, nah.. though kuddos you worked hard to discover it... Rebench eg on XP... and see the difference in subtests... In fact it's not your fault each run is bugged, though as listed above there are discrepancies in your score outcome. Some tests can't score higher than others, but you will discover that quickly when comparing with other scores (see which tests crap out...) Drivers can boost performance, but that doesn't mean ( especially in 3Dmark03 which is very GPU dependant ) that you can surpass guys that run eg 25% faster clocks with decent CPU speeds... and your nature score goes close to ...% faster than anyone elses... doesn't seem logic does it... if you fail to understand ask your dad why that chair is called a chair... he will tell you son, it's like that... I hardly overlook scores, but when I have a similar card and look at the scores and see something weird, then it's normal it will be rectified...

Batches mean close to nothing on SB, in the same batch we found 3 decent CPUs around 5.6 to 5.73... but there was also a few 4.8Ghz ones lol out of 60 CPUs that's pretty crap...

Is that dewar empty now ?

congratz Nick, amasing ( again ) Time to start binning again

Is he the leak ? the one that knows the tools of the trade ?

OCX is all Sergio...

Don't flatter him too much Christian plz

Good revenge from the OC Academay , way to go Luc !!!

Yes Dino Gamers P67A UD5 B3 board has the same C1 state auto/disabled set... some of your bios engineers drink too much.... or... too little Too bad I won't get a bios to test for the P67A UD5 B3, as I used up all my bios credits at GB for the upcoming weeks... looooooooooooooooooooooooool

Got a new bios from GB (excellent job lads), Turbo issues seem to be solved... will ask if I'm allowed to share.... Let's hope they rectify the situation for the UD5 B3 too... (and other boards if needed)

So why do all the other vendors support 1600-1866 and even 2133 with full turbos ? Why limit the Trubo speed of your CPU when you use 1600mhz ? Got a new beta bios to try out, will report back to see if it performs as it should without manual intervention... This one is for all the reviewers that crapped out, sold out and give awards for free !! And test close to nothing lol

So why do all the other vendors support 1600-1866 and even 2133 with full turbos... Got a new beta bios to try out, will report back to see if it performs as it should without manual intervention...

And it's not over yet Did the basic suite, some impressive numbers Asus P67 Dlx vs GB Z68 UD3H Wprime 1024 : 202 secs vs 214 Cinebench : 25126 vs 22802 PcMark05: 14212 vs 12221 Vantage : 20479 vs 19957 Gonna hook up the M4E to the same OS, either it is completely borked now or the Turbo is still somehow somewhere limited... Running OC'ed I get similar numbers than the other boards... it's just at stock that it goes completely nuts...

It seems GB implemented the AUTO bios setting for the C states from 1600Mhz divider to disabled. Only 1066 and 1333 will be enabled if AUTO is selected. This to enhance ram compatibility (they say). The bios engineers spotted the issue quite rapdily ( as they themselves implemented it ofc lol) Solution is to either force it manually or that they change it in the bios... My point is that the normal end user is not required to change this stuff as this should work properly out of the box... enahncing stability okay, but at the cost of 200-300mhz seems an awkward workaround... But thanks Dino for the ahum Live support

Some more testing with the Gskill Flares : seems the only correct working divider is 1333Mhz. 1600-2133 stop at 3500Mhz Conclusion : Dino fails Total Failing scores : Leeghoofd 10 - 1 Dino