another question, for 10.00 more i can get the e762 classified, does it oc as well as the e760?
Printable View
another question, for 10.00 more i can get the e762 classified, does it oc as well as the e760?
I had E760 but it died and was replaced by E762 and I made the same settings and the CPU is working the same way. So they clock the same way. There are diff settings when you want to go for more bclk, but under 220 its the same.Quote:
Originally Posted by Shiranui Gen-An
Yeah it runs very hot. I`m using WC only on the CPU. Running 21x210 with 1.4vCore and 1.4vtt uncore = 4000 and my NB temperature is 60-62(idle load its the same) with ambient temp ~25(case HAF932). When gaming a long time temperature goes +5 degrees cause the card is on the first PCIe slot, but it is a 5850 with accelero S1 and its not a hot one ~55-60 degrees at full load. If u are using a hotter one the NB temps will go crazy + using an air cooling for the CPU...Quote:
Originally Posted by OC Nub
Btw I`m talking about E762, E760 is 10 degrees lower cause there are no 2xNF200 chips. My NB temp sometimes goes over 70 degrees, about 74-5 but that was when I was using 1.2v NB + 30 ambient + long time gaming.. now I`m using 1.15v NB(never tried to lower more running 210bclk for 24/7). According to eVGA there is no problem for the NB running hot as soon as it never goes over 75 degrees(I saw this somewhere on the eVGA forums), and I can confirm that mine is running lower than 60 degrees only in winter :D and never had problems. If u are not running multiGPU the temp u should be worried about is the pwm temp. I`m cooling them with 120mm fan and with 1.4v and without vdroop temp never goes over 60-65, but without fan you ll easily pass 100-120 degrees which is not good. For the NB I`m using only the default 20mm fan coming with E762.
Like I already said they both clock great ;). Using the same CPU the max clock and the 24/7 clock is the same on both E760 and E762.Quote:
Originally Posted by OC Nub
But if I can choose now I`ll go for E760 and that`s not cause it clocks better, that`s because its smaller and can easily fit in any case and its not running as hot as the E762. The only advantage of E762 is if u are planning to use more than 2 VGAs.
thanks for the reply!
my cpu is watercooled, i keep an open case with a couple scythe 133cfm's to help cool the board, also an external house fan. for the extra 10.00 i was think of trying the e762 but only if it will oc as well. also is the pci-e x1 slot functional for my sound card on the e760, it looks aweful close to the top pci-ex16 slot?
Some preliminary testing with the W3540 and my Super Talent DDR3-2000 C7 kit. I always have to use about 1.70v on the memory, none of my processors have an IMC that can handle that tight timing with VDIMM at 1.65v. I had to raise VTT a bit higher than usual in the BIOS but this board undervolts QPI/VTT so it's not a big deal.
http://a.imageshack.us/img843/3707/e...clkstt1.th.jpg
http://a.imageshack.us/img194/3707/e...clkstt1.th.jpg
Actually if you are talking about the voltages shown by the software u better take the DMM and measure it.
Memory voltage 1.70v BIOS = 1.77v real. Classified(760, 762 doesnt matter) is overvolting the memory with 0.07v.
VTT and other voltages I can say that they are close to what u set in BIOS.
As for the clock ability of E762.
3D Mark 06 CPU on WC in the winter.
http://a.imageshack.us/img13/9899/771c686d412c7cfa.jpg
24/7 running with water cooling again.
http://a.imageshack.us/img517/133/i7...lenduncore.jpg
And super pi with CPU on SS.
http://a.imageshack.us/img6/7913/b9f028a0f5fdaad8.jpg
Mine doesn't overvolt VDIMM by anywhere near that much, More like 0.02v measured from the voltage read points.
I measured mine with 1.65v set in BIOS and the actual one was 1.72v. May be when setting higher V, the overvolting is less didnt try to measure all just wanted to know what the real voltage is for my 24/7 run. Or may be its cause of the diff BIOSs I`m using an old one 44 I guess its working good for bloomfield and never tried a newer one.
my board overvolts vdimm 0.02-0.04v :(
but @ least i dont get the droop like on the regular 3x sli board
I tried BIOS 44 after what The Game said. That seemed to get things going. Sad that the newest BIOS isn't always the best, I can only guess it is optimized for Gulftown. At any rate I was able to do 19x210 DDR3-2100 CAS8 HyperPi 32M stable. I used a different kit this time, my Dominator GT 1866 C7 Elpida-Hyper based and found they needed more VDIMM than my Super Talent sticks. The Everest numbers and HyperPi times aren't good because I loosened tFAW and B2B CAS delay probably more than I should have but I was just trying to get 8x32M to pass.
http://a.imageshack.us/img153/4858/e...clkdomi.th.jpg
http://a.imageshack.us/img818/4858/e...clkdomi.th.jpg
http://a.imageshack.us/img641/4858/e...clkdomi.th.jpg
What about this?
If I use gnd from EVGA EZ Voltages points, I get 1.65v (I set 1.60v in BIOS), but if I use ANY other gnd (from molex, atx connector etc.) I get like +0.02v so in this case it is 1.67v. What is correct voltage?
That`s what I heard the newer BIOSes are gulftown optimized its normal. Usually I hate to try every new BIOS and test again if my 24/7 run is still stable. When I found a good one, for 24/7 and bench I stopped updating and I`m glad I didnt miss anything sticking with this one :D.Quote:
Originally Posted by Shiranui Gen-An
Hmm, I always use molex for GND, its way easier and I get 0.07v diffrence 1.65v BIOS -> 1.72v DMM. Is there a diffrence with any other voltage or only the memory?Quote:
Originally Posted by quake6
All voltages are like +0.02v higher with any other GND except on evga ez points. I get the same value with evga ez voltage gnd on voltimeter and in bios monitor but with gnd from molex it is +0.02v higher on voltmeter then in bios monitor
Hi guys ~
For a long time with my E760
Still shows good performance.:)
E760 / 980x air-cooling super-pi 32m test
Using the dual channel memory
http://cfile23.uf.tistory.com/origin...4C56141A729AC8
▲ i7-980X @4G DDR3 2000 7-9-7-24 1T
http://cfile29.uf.tistory.com/origin...4C56141B737A88
▲ i7-980X @4G DDR3 2000 7-7-6-21 1T
http://cfile25.uf.tistory.com/origin...4C561419715BCD
▲ i7-980X @4.6G DDR3 2000 7-9-7-24 1T
http://cfile8.uf.tistory.com/image/1...4C56141C74DCB7
i'd go with the 760Quote:
Originally Posted by OC Nub
i have the p6t7 supercomputer with the nf200 chips get hot
i honestly dont care for all 16x pcie
but it was coolt to run 3 gtx 285 tri sli+ a 8800gt for physx just for kicks
i still plan on watercoolin the nb but with fans on nb and pwn runs good
nice clocks namegt
Shiranui Gen-An dont go nuts on namegt's clocks its way too easy to clock memory with 980x ;) i sold my 980x cause was too easy n got bored
My EVGA 760 Classfied Photos :)
Memory full
http://cfile29.uf.tistory.com/image/...4C4F919BAC6B36
http://cfile25.uf.tistory.com/image/...4C4F91A0B55D4E
http://cfile29.uf.tistory.com/image/...4C4F91A2B95477
http://cfile4.uf.tistory.com/image/1...4C4F91A1B742EA
http://cfile21.uf.tistory.com/image/...4C4F91A3BC793B
Nice shots NAMEGT...can you please tell us a little more about that Geil memory?? Works fine on eVGA board??
Quote:
Originally Posted by Expat GriZ
EVGA board and the beautiful harmony;)
Other operating speed of the memory.
OC a little bit bad
However, [1600Mhz cl7] cl9 memory timings and 2000MHz memory can be
DDR3 GAMING EVO TWO TRIPLE CHANNEL
PC3-12800 1600MHz C7 EVO TWO TRIPLE CHANNEL
DDR3 GAMING EVO TWO TRIPLE CHANNEL
PC3-16000 2000MHz C9 EVO TWO TRIPLE CHANNEL
http://cfile6.uf.tistory.com/image/1...4C4F919AA83EFA
http://cfile29.uf.tistory.com/image/...4C4F919BAACBDB
http://cfile6.uf.tistory.com/image/1...4C4F919BABD6C3
http://cfile29.uf.tistory.com/image/...4C4F919FB3D81E
My board overvolting the memory around 0.3v (rev.13 e760)Quote:
Originally Posted by The_Game
How did you measure it?
I always use DMM ;)
The difference in software is 0.2v
That's exactly why I want a Gulftown chip, I can see what the memory can do without uncore holding them back :rofl:Quote:
Originally Posted by zsamz_
its not just uncoreQuote:
Originally Posted by Shiranui Gen-An
the memory controler is real good
but a waste of $
i had a 980x for a month and got bored with it ,there is no challenge its like driving on auto
If you wanted a challenge you should have traded your 980x for a few of my chips :ROTF:Quote:
Originally Posted by zsamz_
all your chips are hexed:ROTF:Quote:
Originally Posted by Shiranui Gen-An
even if you get a 980x it be hexed:ROTF:
I repeated some of my previous clocks with this setup but was able to lower voltages somewhat. I'm not sure if that's due to the Classified just being better or due to me being willing to tweak more.
23x175BCLK 1.2625v DDR3-2100 8-8-8-24 2T 1.75v
http://a.imageshack.us/img46/8866/e7...lkdomid.th.jpg
http://a.imageshack.us/img444/4876/e...clkdomi.th.jpg
http://a.imageshack.us/img46/4876/e7...clkdomi.th.jpg
http://a.imageshack.us/img194/4876/e...clkdomi.th.jpg
21x210BCLK 1.325v DDR3-2000 8-8-8-20 1T 1.65v
http://a.imageshack.us/img62/4199/e7...clkdomi.th.jpg
http://a.imageshack.us/img180/4199/e...clkdomi.th.jpg
http://a.imageshack.us/img802/4199/e...clkdomi.th.jpg
i can tell you for sure its not your tweaking lolQuote:
Originally Posted by Shiranui Gen-An
This is getting a bit silly. I tweaked volts some more and only needed 1.425v VTT for 21x200 DDR3-2000 CL7. Still need more than 1.70v for the Dominator GTs to do CAS7 but hey, there has to be some challenge right?
http://a.imageshack.us/img248/4199/e...clkdomi.th.jpg
http://a.imageshack.us/img80/4199/e7...clkdomi.th.jpg
http://a.imageshack.us/img827/4199/e...clkdomi.th.jpg
Forum crashed while I was posting and did a double post.
Did anyone experience an oddity in that their 1 year old stable overclock all of a sudden is no longer stable?
I noticed this when I would get black screens (monitor goes into standby, game music keeps playing and shortly after a hard lock)
happened first with SC2 then the FFXIV benchmark and now prime is no longer stable for more then 2 hours.
I'm wondering what could be giving up after time?
May be the hot summer days are causing the problems? That`s why when I`m testing my 24/7 overclock in the winter at 20-22 ambient I`m testing with prime or linx. I know that if I`m close to 80-85 degrees it wont be prime/linx stable again in summer days when the ambient temp is 30+ but I dont care cause I cant really load the CPU as much as these 2, and even now max CPU temps I can see for normal work not benches are 60-70 degrees. Yeah it happens to me 12 hours prime95 to be unstable after a period of time but I know the reason(overheating) so I just dont run prime in the summer :D. Also never had problems in games or whatever as soon as the CPU is not overheating. Anyway 2 hours prime should be stable enough for gaming. I think its not ur CPU may be the PWMs or NB are overheating or may be the memory. I have played games so many times with CPU which cant hold more than 10 sec prime and never had problems :D, but I guess its accoriding to the game or the CPU.
Same Bios?Quote:
Originally Posted by Bojamijams
I am still thinking about trying one of these boards. Was looking at the e760 classified, but waited and now it is more $$$ than the e762. Can get the e762 for 389.00 at newegg, may just go for that. Only concern I have is still the board temps, have no plans to buy the block for the board and change my loop.
Yeah same bios, same everything. Not really convinced its the heat as my system is watercooled and has far more heat dissipation capable then anything can put it under.
I've heard of the effect called 'electron strain' where the silicon starts leaking current due to being under extra voltage then 'normal' for some time. In this case usually its required to increase the vcore a bit to maintain the same overclock but I've tried that and still nothing. I'm wondering if RAM can start giving out after a while though I haven't ever fed it more then 1.65V
You change gfx cards, or clocks on the cards? I got the black screen on my 5870 last night playing bc2 on a tested stable oc with my cpu, but gpu at 950 core needed more juice.
Did change the gfx cards but its possible that the new FFXIV benchmark (which is the quickest way for me to test stability) didn't like my previous clocks which were stable under Vantage and any game I've put it under. I've had to backdown from 1025core to 950core. Trying 975 produced the same error. This is maxed at 1.3VGpu.
Problem may have been two fold though. Gonna have to find Linpack again and see if the CPU/RAM connection is indeed fine.
Doesnt look like the e762 will even come close to fitting in my antec 1200 case lol. newegg has it for 399.00 with a 100.00 mir, not bad 299.00 for that board though.
Sweet deal, I paid $299 for mine off eBay.Quote:
Originally Posted by OC Nub
Hey guys just setup my e760 and it seems to be working great.. I am having one issue.
The machine is in standby mode (blue light) and is reporting error "FF" does anyone know wtf this means?
damn just got a second 760 lolQuote:
Originally Posted by OC Nub
i need to sell some x58 boards lol
That "error" means: full function. :ROTF:Quote:
Originally Posted by Kenjiwing
Then why do I have a blue light indicating standby @_@Quote:
Originally Posted by SAE
i7 920 D0 @ 4.2Ghz
EVGA X58 3X SLI Classified E760
Corsair Dominator GT CMG6GX3M3A2000C7 6GB @ 1600Mhz CL6-6-6-18-1T
uncore 3.2Ghz
http://img571.imageshack.us/img571/4...ncore32ghz.jpg
uncore 3.6Ghz
http://img686.imageshack.us/img686/8...ncore36ghz.jpg
uncore 4.0Ghz
http://img693.imageshack.us/img693/6...uncore4ghz.jpg
uncore 4.2Ghz
http://img228.imageshack.us/img228/5...ncore42ghz.jpg
stop here as core = uncore.
LOL My CPU can also run uncore 4000 with 1.4vtt, but there is no chance to stabilize 4200, it can boot and bench with 1.4vtt again, but even 1.5v is not helping for stable work.
With higher uncore do you see better perfomance in gaming/folding, or is it just for benching? Is vtt the only voltage that needs to be adjusted for higher uncore?
my too :)Quote:
Originally Posted by The_Game
This is not a big deal with mem rated 2:8 700-900Mhz , my problems starts from 2:10 1000Mhz + :DQuote:
Originally Posted by owikh84
Quote:
Originally Posted by The_Game
Increasing Vdimm might help (as well). Stability also depends on the right relation of vdimm/vtt.Quote:
Originally Posted by ChaosAD
straight boot to windows:D
there is hope :up:
now gotta get her stable
n then push more
great start man, maybe you can top your rexII linx run.:up:Quote:
Originally Posted by zsamz_
thats the plan;)Quote:
Originally Posted by OC Nub
i spent almost 2k :rofl: just for that i shoulda just got a phase unit lol
lol, think you had more fun testing all your boards though. (except the rex3)Quote:
Originally Posted by zsamz_
this weekend im taking the bad clocking 960 and put it in the rex3 again and throw some volts at it and see what it will do, not a huge loss if they both die really.
sweet.cant wait to see how far you can go with the E760Quote:
Originally Posted by zsamz_
Thats sick zsamz. Great clock. Which bios did you use for this? Can you post voltages that you used? What cooling and temps? Post screens :) Wish you the best for your 4.8Ghz linx run :up:
once i play with it a bit more i will post a full bios templateQuote:
Originally Posted by ChaosAD
i was using bios 664
cooling is 1 thermochill 120.3 on portable a/c
i'm gonna redo loop soon and put in another rad + nb block or a full block
You dont really need nb block, one extra fan is enough imo
what temps you have ?
btw: wish you luck with linX ;)
I'd suggest you do the Full cover block it keeps those pesky VRM temps down big time.Quote:
Originally Posted by zsamz_
Yep go with a full block. Vrm temps are the issue not NB temps. TC 120.3 is good, but ac unit is even better :up: Not like my room, my poor pc has to deal with 30C ambient :shakes:
Edit: I guess u mean bios 64, not 664 right? :)
yes 64Quote:
Originally Posted by ChaosAD
i'm mainly concerned with vrm temps and fan noise
Is bios 64 considered the better/best bios for an i7 920?
I find new releases worse than previous bioses, new releases are optimized for Gulftown but i might be wrong :)
This one is the best for my cpu
http://img405.imageshack.us/img405/3...wanielc.th.jpg
Uploaded with ImageShack.us
I've been playing about with lowering the PLL voltages (my Asus didn't have such a useful feature) and so far lowering both CPU and IOH PLL from 1.8v stock to 1.425v has allowed me to lower my vcore 4 notches in the bios, from 1.35v down to 1.325v and remain (thus far) totally stable.
Yeah, usually lowering PLL is helping to lower the Vcore, but sometimes there are problems like its stable at stability tests, but there are BSODs when idle, so you ll need more testing than linx/prime to find if its stable :).
Folding 24/7 with the SMP BigAdv Client does the job nicely. LinX/Prime don't have anything on the ability Folding has to bring an overclock to it's knees. :D
On my bloomfield 970 I actually achieved better clocks. All in all the newer bioses just have more options. Been testing my 980X but I need to do a psu swap to my silverstone 1500 watt my 1250 needs an RMA cause my 3.3v in dropping bad.Quote:
Originally Posted by 71proste
How better clocks, whats the difference ? What bios version ?
about 100mhz more, but I'm using 6 sticks. bios versions 64 and up.Quote:
Originally Posted by 71proste
Check out the evga forums for details on the differences.
Quote:
Originally Posted by http://www.evga.com/support/drivers/default.asp?switch=2
I am having the oddest problem with my E762 Classified. The board simply refuses to clock 3 out of my 4 Elpida Hyper-based kits. I have 3x2GB Super Talent 2000 C7, 3x2GB Super Talent 2200 C8, 3x2GB Patriot Viper II 2000 C8 and 3x2GB Dominator GT 1866 C7. The Corsair kit is the only one it will do anything with, the other kits I get freezes, WIndows not booting, etc even using settings that worked for the Dominators. I'm sure it's not the memory since they all clocked fine in my Asus P6X58D Premium. I've used 3 different CPUs as well. Any tips? I've played with vcore, vtt, vdimm, to no avail. It's very puzzling.
i'm puzzled lolQuote:
Originally Posted by Shiranui Gen-An
be more precise this is not like making bread lol
i can make all my hypers dance on the board :D
i didnt go over 2200 yet though
Just an example, DDR3-2000 7-8-7-20 1T on the 2:10 divider (200BCLK). With about 1.40v set for VTT and 1.70v for memory the Dominators will pass 8x HyperPi 32M happily. The Super Talents will freeze generally before loop 1 even finishes. Changing VDIMM and VTT and even vcore doesn't help, and I know the sticks can do it because they passed at those settings on the Asus. Dropping to 2T doesn't help, I've also tweaked RTL, tFAW to no avail. The Super Talent sticks use MGH-E just like the Dominators, while the Patriots use MNH-E. However the common thread between the Patriots and the Super Talent sticks, the ones that are giving problems, is the type of PCB, 8-layer Brain Power. The Dominators seem to use their own Corsair-made PCB.
some ram just need a little more qpi volts man that simpleQuote:
Originally Posted by Shiranui Gen-An
Finally managed DDR3-2100 C8 with the Super Talent 2200 C8 kit. Sadly I still need lots of VDIMM but oddly enough high VTT actually made things unstable; I thought I'd need 1.55V+ to pass 8x HyperPi 32M but 1.52V was what did the job.
http://a.imageshack.us/img825/7889/e...x210bcl.th.jpg
:yepp: with me 1.55vttmakes it unstableQuote:
Originally Posted by Shiranui Gen-An
but thats alot of juice for vdimm
It's got be something else Gen-An. My 920 isn't particularly strong. Have you tried bumping volts elsewhere like QPI PLL or IOH Vcore. Also as I've mentioned before, loosening write to write and read to read same rank did help with my sticks.
Bios 64
HyperPi 4.2 ghz mem 2000mhz 7-8-7-2T @ 1.67v
http://img820.imageshack.us/img820/8...8742167.th.jpg
HyperPi 4.2 ghz mem 2000mhz 7-7-7-1T @ 1.68v
http://img52.imageshack.us/img52/840...742168b.th.jpg
RLM
BIOS used makes a huge difference, I wasn't able to get anywhere on 64 or newer. Went back to 44 for that last run.
Okay I'm totally puzzled now. I hadn't been able to get these STT sticks to clock worth anything on my Classy lately, I mean simple stuff like DDR3-2000 7-8-7. I was getting freezes as soon as HyperPi 32M would start up. I tried several BIOS revisions along with tweaking VTT, VDIMM, vcore, tFAW, RTL, you name it. Finally I went back to BIOS 44 and then I managed 2100 C8 but VDIMM was still higher than I liked. Today I kept the same settings that had passed at 1.74v memory but lowered to 1.66v and to my utter surprise, it passed 8x HyperPi 32M.
http://a.imageshack.us/img815/7889/e...x210bcl.th.jpg
http://a.imageshack.us/img842/7889/e...x210bcl.th.jpg
1.66V are more like real 1.69V (DMM). Still too much for my tastes. Are those the CL7 from your sig? If so, I'd contact SST asking for known problems with evga (asus worked fine?).
then your cpu like one of mine it dont like too much vdimmQuote:
Originally Posted by Shiranui Gen-An
Pissing against the wind there. I tried that, and they quickly blamed the classified. Their support isn't great.Quote:
1.66V are more like real 1.69V (DMM). Still too much for my tastes. Are those the CL7 from your sig? If so, I'd contact SST asking for known problems with evga (asus worked fine?).
RLM
The 2200 C8 actually. To date I haven't been able to run 2000 CL7 or 2100 CL8 without setting BIOS volts to 1.68+ on any board.Quote:
Originally Posted by SAE
Wow, that (they) sucks. :down:
I just picked up a EVGA Classified 4way SLI and was wondering if i have to flash the BIOS to the latest version before installing a 32nm westmere cpu?
You need to check the bios version on it and the first bios to support 32nm...
I've noticed something else that the lower CPU and IOH vPLL has fixed. My cores are all equal temperatures now. Two @ 66c LOAD and two @ 59-60 LOAD. Before it was 75+, 70, 72, 55 or there abouts @ LOAD. I tried with a lapped TRUE, a Megahalems, a Swifty GTZ and finally a Swifty XT and nothing evened out the cores, not even lapping the i7 IHS. Dropped the above PLL and BAM, cores are even. :D
I'm still totally baffled by this board when paired with my Super Talent sticks. I managed 2100 C8 HyperPi 32M stable but just trying 2000 C8 last night it froze. ?!?!?
Curious if i am not aware
Do u guys experience write fail (red block) during bios flashing? They are block #1,2,4 & 5.
But successfully flashed the bios until finish
nope but i would reflash the bios if i were youQuote:
Originally Posted by owikh84
reflashed to older bios, same 4 red blocks, still write failQuote:
Originally Posted by zsamz_
I tried to flash my E760 and I got cmos error during reboot. Then when I restart the computer the error msg goes away but only comes back on if I change CPU...
:shakes:
Could it be a faulty bios chip for my case.Quote:
Originally Posted by kill_a_wat
Altho write fail at the 4 blocks my system still acts normally.
Tried 74 and 73 same outcome. hmm
red blocks but flashing finished till the end and press F1 to reboot normally
btw im using USB thumbdrive flashing method but dont think other methods will make different right
EDIT:problem fixed
used command /sn/py/ instead of /cks/sn/py/cc/cd/cp/ld/WB/ch
I used CD method, but I don't think it matters on the method. Just the end result sucks :(:shakes:
Switched out the STT sticks for a Dominator GT 2000 C8 kit. Lookie here.
http://a.imageshack.us/img834/865/e7...clkdomi.th.jpg
http://a.imageshack.us/img837/865/e7...clkdomi.th.jpg
http://a.imageshack.us/img832/865/e7...clkdomi.th.jpg
looks like you finally got a good imc:up:Quote:
Originally Posted by Shiranui Gen-An
but try usin 21 multi its better
hi guys, i'm testing an EVGA 760 but i've some weird problems.
using liquid cooling on the CPUs the system can reach 225 MHZ of BCLK with 920 and 210/212 with the 950s.
I tried different rams (kingston 2000 cas8, kingston 2133 cas9 and ocz platinum 1600 cas7) without luck.
i've also tried many bios, the best so far seems to be the E76044.
at first time i though "here is damn hot, maybe that's the problem", then i fired up my single stage phase change.
for the first test i used my i7 920 (capable of 255 mhz of bclk on other board), with QPI in full mode i stuck near 225 even at -50° C, post code error display show 89 error.
with QPI in slow mode i can boot at 245 but i still have some stability issue (the cpu can do super-pi 32M at 245 on other board, then she isn't the problem).
with the 950s at -50° C i can reach 220 MHZ of BCLK with QPI in full mode, when i set it in slow mode i can't ever see the bios, the post code error display going crazy ad show each time different errors.
another problem is that when i'm using H2O i can use the turbo (21x for 920 and 24x for 950s), when using the single stage there is no way to get it work.
i've tried exactly the same settings that i use at H2O but turbo still wont work at -50° C.
using the X-COOL jumper near the SB doesn't change nothing.
this mobo is driving me mad....do you have any tips guys? i'm starting to thinking that the mobo is half dead and that's strange because i bought it brand new from EVGA europe site one week ago.
mafio try reflashing bios might be corrupt
thanks for the quick reply mate :)Quote:
Originally Posted by zsamz_
what string did you use to flash?
i use this: AWDFLASH biosName.bin /sn /cc /cd /wb
i burned cdQuote:
Originally Posted by Mafio
ah, ok, then you use CD method, i'm using USB like i did with all my DFI.Quote:
Originally Posted by zsamz_
using AWDFLASH biosName.BIN /cks/sn/py/cc/cd/cp/ld/WB/ch awdflash show that the block 1, 2, 4 and 5 aren't writable.
that's might be the problem.
i'm going to burn a cd hoping that this time the board start to work like she should work......
flash with cd till you get proper writeQuote:
Originally Posted by Mafio
The first 5 blocks are never writable Mafio, i don't know why, but it never causes a problem. The reason you may be "seeing" it as a problem is because you are using switches when you flash. You don't need to do that at all.Quote:
Originally Posted by Mafio
Just make shure the USB stick is bootable, put AWD Flash on it and the bios that you want to flash to. Boot to the USB stick, type awdflash and the utility will appear on the screen, then type in the Bin file that you want to flash too. It should then just flash as normal.
cd method don't update the first 20/25 blocks and unfortunately the corrupted/unwritable blocks are the #1, 2, 4 and 5.
edit: kitfit1 i flashed like you said without using autoexe.bat but nothing changed.
i can do everything but there is no f(l)ucking way to boot over about 220 MHZ of BCLK, and i don't think that i did something wrong.
probably i'll go for rma.
thanks for the help guys
I was just issued an RMA by STT, I don't think I'll even test the replacements when they get here.Quote:
Originally Posted by RLM
My 4-Way is pretty spot on to what is set in the BIOS for VDIMM on load, I always use the ground point on the board.Quote:
Originally Posted by SAE