Printable View
Can someone please explain what disabling the "Keep DMI Data" does in Q-Flash. I have tried to find an answer but am still puzzled. After the data is cleared is it rebuilt when booting into Windows. Are there any possible drawbacks or is there any risk in clearing this data when flashing. I am just nervous I guess because I haven't used it before and because it is not the default setting. Thanks.
Quote:
Originally Posted by Falkentyne
Sorry Falkentyne and Kent10, I was asleep at the wheel when I mentioned that! I was thinking of the Marvell firmware update, the Intel ME does not show on screen like I mentioned, sorry for the confusion!Quote:
Originally Posted by Kent10
Quote:
Originally Posted by Kent10
Nice to see you found a working USB stick for Qflash! Some work, some just will not, and the best chances for compatibility are with smaller ones like 2GB or less, some larger ones will work but best chances are smaller ones.Quote:
Originally Posted by Kent10
Clearing DMI deletes all your old settings, this should always be done so that no older settings are carried over. Look at each new BIOS like it's your first day with the board, with each new BIOS you'll need to tinker and tweak settings until you are stable anyway, so no need to use old settings anyway as they might hold you back.
I've always had my settings properly cleared, without using the /C clear dmi, just by powering off after the flash (without rebooting), turning off the PSU, and then holding down the clear cmos button for 30 seconds. That's worked every time.
It normally does clear them, but not always, and sometimes not hidden or removed settings. And ya, a clear CMOS after flash would clear it out too!
I went ahead and tried F7c today. This is what I found.
First, I did everything everyone suggested. Used Q-flash, cleared the DMI, loaded optimized defaults before and after flash and rebooted both times. There were no problems with flashing. The only thing I didn't do yet was flash the backup. I thought I would try that later because I wanted to test stability first. I remembered what Lsd said about having to tweak my BIOS settings for each new BIOS. First I tried the same settings I used with F4 and ran P95. I blue screened after 10 minutes. Either my processor has degraded since I tested it under F4 with several hours of P95 or the BIOS is different enough that it is doing something else to cause instability. The voltage readings in HWiNFO were the same for each BIOS if that means anything.
The only reason I wanted to try a newer BIOS was in case I wanted to upgrade to Ivy Bridge. I think I will stick with F4 for now since it is working for me and I don't want to have to increase my VCORE more for F7c or decrease my OC. When Ivy Bridge comes out I will decide if it is worth the upgrade and if I get it then I will have to try the new BIOS with it.
Thanks everyone.
I grabbed 2 kits of this memory http://www.newegg.com/Product/Produc...82E16820233258 to try with x79, and p67. Last night i installed 1 kit/2 sticks on the P67A UD5, it will not boot at 2133mhz no matter how high I raise vdimm. I am posting this at 2333 on the r4e x79 board, so I am pretty sure its not the memory. They are dual channel kits and should be compatible with this platform.
Tried switching slots, and even xmp profile, made sure I am on the latest bios, but no luck.
Any suggestions on what the problem my be? Should I hold my breath on a bios update or just accept that they will not be compatable with the Gigabyte 1155 boards?
Well I have a lot of memory that sucks on LGA1155 but is amazing on X79. I have two kits of XMS 2000mhz cas 9 i think, i bought two dual channel kits. They run at 2000mhz on the Z68 boards I have, but with a lot of work, however they run at 2400 on X79 without much effort.
Do you have another LGA1155 board to try it out on?
BTW I have that kit, but the 4GB version, it works at 2133 on my UD7, i am pretty sure. i haven't tried it out for a while. is your kit ver a or b?
Yeah lol, now I have a lot of memory that sucks also. I think its the memory, I finally got the Z68XP-UD5 to run in dual channel again, and this memory also sucks on this board as well.
On NewEggs site the memory is rated 2133 9-11-10-27 1.5v, and on the actual memory it reads 2133 9-11-9-27 1.5v, it wont do either unless I use XMP profile which sets 9-11-10 2t 1.5v with trfc at like 178 lol. It is pretty much garbage, especially considering what the are charging.
For anyone with startup boot loop or other issues, here's another official update guys. I sent in a report on a few users still seeming to have these issues over the past two weeks, so here is there testing results and full reply.
Please pass this info along to your fellow Gigabyte users, thanks!
So be careful with that reset button and it should also work properly now too! But I still say avoid if/when possible :DQuote:
Test 3 models of Z68 boards with the following BIOS:
1.) Z68X-UD7 with BIOS F10
2.) Z68XP-UD3 with BIOS F9g
3.) Z68AP-DS3 with BIOS FB
Our test results:
1.) Press CTRL + ALT + DEL keys when during boot up or in DOS mode >> Cannot see “continuous reboot loop”, nor “Waiting for ME ready”, nor “system overclocking failure” message. The system just reboot fine.
2.) Press the Reset button
A.) Press the button quickly and release immediately >> Cannot see any issue.
B.) Press the reset button for one second or so (This is not normal behavior.) >> System will shut down, and then power on by itself. Then the “Waiting for ME ready” message appears and then system will shut down and power on by itself again. Sometimes, it will show “System overclock failure” message.
Moreover…
According our testing before,
the issue of “continuous reboot loop” can be fixed by flashing the Backup BIOS with the same version as the main BIOS (latest one).
We Flash the Backup BIOS with the same BIOS version for the main BIOS version before test.
And always keep backup BIOS up to date matching with the main BIOS.
Final BIOS out for P67/Z68 boards.
I am going to toss the Z68X-UD7 on the bench this weekend, I will try out that new Final BIOS then, I tried the beta and it fixed the boot loop. However LLC still needs tweaking.
This is just bad LLC :(
F8 Final, P67 UD5 B3,
1.415v Bios, LLC7, when playing a single core league of legends...(prime was perfectly predictable at 1.416v small FFT, 1.428v blend).
http://img4.imageshack.us/img4/5593/seeo.jpg
this is in a SINGLE threaded game!!!!
They REALLY need to bring back the old LLC2.
I tested the old LLC2 extensively...on my P67 UD5, the vcore would have been 1.428v in idle 1.440v in league of legends, and 1.452v MAX at heavy load (with Bios set to 1.415v). MUCH better than this...(as long as you watch out for idle BSOD you're fine...)
^ I always used LLC1, never cared for LLC2, and never understood why people wanted vcore to increase under load.
I do agree they should go back to using LLC1 and LLC2, wtf do we need 10 levels for? 2 or 3 levels that work correctly would be perfect.
Latest bios seems to be working very well. System seems to be very stable, and recovery from failed oc is as good as any board Ive owned.
I tried the Maximus 4 Extreme this weekend, wasnt as impressed with it as I had hoped, so Im now back on the Z68XP-UD5. I take back some of my :banana::banana::banana::banana::banana:ing and complaining about Gigabyte P67/Z68 boards, newer bios seems to really make a diference. However it would have been nice to not have to wait so long.
Great to hear that OC Nub, I'm glad to see the new BIOSes are working better for you now!!
I cannot begin to tell you how much I agree with this.Quote:
Originally Posted by OC Nub
I try to figure if a setup i have done on the P67 UD7 could be the reason who have kill one of my HD7970 ..
As you know 7970 have the "zerocore " features. who completely shut down the second card when not used ( CFX or if the monitor is shutted of ).
- first slot: HD7970 16x ( slot 1 )
- Second : HD7970 16x ( slot 2 )
I was run my Soundblaster Titanium HD then in the last PCI express slots... given me 16x - 8x and 1x for the soundblaster.
The other day, i have remove the Titanium and tried put the second 7970 in a slot far of the first. ( for get more space between the 2 cards, cards was run on 16x and 8x ).
2 hours then, i go make the dinner, let the PC run... When i come back, something was going wrong with FF. i reboot the PC; the fan of the second card is running at 100% speed, but the card is not recognised under windows, not even seen. Card is in RMA, so it is not a big problem.
Any idea ? can it be the switch + Zerocore feature ? or something else ? Can be just the card who had a problem anyway
you place the sound blaster in the last 16x slot?
BTW that zerocore features sucks when trying to bench sli...
last pci express slot ( who is 8x not 16x, but who whas working at 1x so )... As it is impossible to use the standard pciexpress 1x, i have use one of the PCIexpress 16x-8x slots for the sound card. I have the feeling this have maybe contributed to kill the card.
For give you an example, if i use Crossfire ( 2x 7970 ) + the Titanium HD installed ( in the last slots ), BF3 and Skyrim had sound problem, the sound was crash at the moment i enter the game. if i use only 1 7970 + Titanium HD ( in any slot ), no problem. Ofc the 2x 7970 + Onboard sound, or a PCI sound card, no problem too.
Its maybe not related, i was ask me if the board who use Nf200 switch, + the fact to use the sound card in one PCI express slot ( normally dedicated for GPU, and who should not be used at 1x ), have contributed to it.
Between the idle states who reduce the PCI express power draw, the Zerocore who shut down completely the card, and the NF200 who is used for the switch on the PCI epxress lanes.. this have maybe do a strange mixtures.
Did you try out the problem card in another slot to rule out that you didnt have a bad pci-e slot?
Quote:
Originally Posted by Lanek
yes ofc, i have tried the 3 other slots, ( including the one where my other 7970 is running right now ) + in another PC ... I wait the return, so no worry, i will get it thoses next days. Just i was ask what do you think about all thoses pci express sates. ( just for dont repeat it, if you think there's a slight possibility it can be related ). New GPU's use a lot of new features for reduce Powerdraw, including on the PCI express slots, as this board is one of the only who use the NF200 for get 16x / 16x or 16x - 8x - 8x. .Quote:
Originally Posted by davidm71
But looking at the response, i will now bet this is surely just the gpu who had a little hardware problem. ( the card was one of the first Sapphire who have been delivered )
Make sure USB 3 Turbo is turned OFF if you're using two crossfired cards with a x1 sound cad. You can also try disabling esata to test. (after you get a RMA replacement card).
Turbo is off, but i will look for the esata when it come back ... Thanks Falken.
You're welcome.. Might not fix anything, but I've seen the PCIE link go from x16 to x8 on my UD5, if USB3 turbo is enabled, since USB3 speed boosts take up a lane. And eSATA also takes up a lane too. Another thing you can try is setting your BCLK to 100.02 instead of default (99.8).
on my UD7 (P67), the bclock ( auto ) is allways at 100.3mhz .. ( with all bios )
I still use the F6, i have find the F7 unstable without LLC enabled, with the F6 i can leave it disabled at 4.5-4.8 and need LLC 6 at 5ghz and 7 at 5.2ghz+ , but with the F7, impossible to let it disable even at 4.5ghz,.. BSOD when windows boot... ( related to my cpu surely ).
Voltage 1.42 in bios for 5ghz ( idle state )
http://img713.imageshack.us/img713/3291/cpug.png