Can anybody show what performance gain you get by overclocking the memory?
Intel specs show Ivy Bridge CPUs are only good to max DDR3-1600. Samsung M379B5273DH0-YK0 shows sweet spot on my UD5H is 8-8-8-22 /1600 with trivial perofrmance gains above 1600.
I simply prefer disabling in bios any unused items, such as the secondary Atheros ethernet port and the firewire controller.Quote:
Originally Posted by sin0822
Whenever a discrete graphics card gets installed, I would then disable the onboard Intel HD 4000 graphics.
Beta bios version F9o works fine, so something new was introduced starting with F9q and later.
Yea I am aware it is less devices pinging the CPU it is helpful for extreme overclocking in some cases but in others doesn't make a difference, like I got 116.95mhz bclk and i didn't disable anything, not even TB. I will put in a request to check it out.Quote:
Originally Posted by vailr
Performance gains from OCed memory are only really noticeable in benchmarks. There really is very little real world advantage, but it is fun and allows you to push things further,t hat is why people OC. Density in memory is more important for daily use than speed.Quote:
Originally Posted by cab
Thank you. When don't they show density in the part # then?Quote:
Originally Posted by sin0822
I have Samsung M379B5273DH0-YK0 and g.skill F3-1600C8Q-16GAB -- which is denser?
Hi guys jus checkin i did things right.
Backup Bios = F8 (OC setting allready set)
Main bios flashed thru Q-flash in bios to F10x not @-Bios (should i have done?)
--- Things i disable when i OC.. Marvel ports, Audio, Wake on LAN, firewire, C1/C3
--- Enable or disable Intel Virtualization tech. (Does this help in any way?)
Everthin flashed fine...rebooted..I always press ESC to go into Advanced bios then F7 to load defaults..Enter...then F10 to Save n Exit..reboots shuts down once as allways then checked bios version F10x correctly displayed. Im sure Dram was showing 1.35V i then booted into windows...
Shutdown...CLR onboard CMOS
Enter Bios Dram showed 1.66 or 1.65 v which i think has allready been logged. I then went on to change to my OC settings. I set dram to 1.5v 2133 10 10 10 28 & 30 1N set VTT IMC (1.095v 1.1v) also tried (1.145 1.150)
Got the deadly beeps lots of boot loops. Then i noticed my dram voltage stuck on 1.47v (this is setting from backup bios) no matter what i select the voltage never changed. I did try 2 x 4GB 2133 which worked (tried HCI memtest no errors) 4 sticks givs me errors in 2 secs
Is F10x for Sammy + IVY only?? should it work with SB aswell ??
I know my max wud be DDR 2133 id be happy with that. Currently on 1866 9 9 9 27 1n no probs.
Need to check the following::
Does both Backup & Main have to be on F10x for the sammy fix to work??
Is Q-Flash the best way to flash ? (ive never had issues with this method)
Voltage for DDR 2133 10 10 10 28 1N is 1.5V ok ??
Any help wud be greatly appreciated.
AFAIK, there's no "crosstalk" at all between the main and backup bioses.Quote:
Originally Posted by irfs
Q-Flash can sometimes fail, such as when upgrading a Z68 board from a standard bios to a UEFI bios.
I normally do not use Q-Flash, and instead use the bootable USB thumb drive method, where the DOS version bios flasher is used.
[The DOS version flasher "efiflash.exe" is included with the F8 bios from the official support site:
http://www.gigabyte.us/products/prod...?pid=4139#bios ]
But that can also fail sometimes, and the Windows beta version B12.0723.1 atbios program did finally work for me, flashing from F10 to F10x.
Had to boot to Windows XP 32-bit to get it (atbios) to install, however.
I have @bios installed allready on Win7 X64 SP1Quote:
Originally Posted by vailr
Are you running 8GB 4GB x 2 config F10x bios? 2133 + should work fine. For me its 4 x 4 16GB + SB cant get to work on 2133
There shouldnt be any crosstalk between bioses but ive seen it happen quite a few times (before Statio advised me to reset CMOS on F10) Cmos reset fixed the Dram voltage for me on F10.
^flashing the bios from Windows is usually a bad idea, and should be a last resort if nothing else works.
Sandy Bridge usually can't use higher multipliers than 21.33x
Also if vDRAm is 1.65v that is because the DRAm multiplier is higher than 24x i was just told, it will auto set that vdimm, just for testing sake.
When i said memory density i just mean the size of the module, like 2gb, 4gb, 8gb.
Also memory that is double sided performs better than single sided in most cases.
Also there IS cross talk between the main and backup BIOS on every boot up, that is one of the good and bad parts of the dual BIOS system. It can cause odd behavior of both BIOSes don't have the same exact settings, like it is one reason BIOS profiles wont transfer after a BIOS flash.
Sin are you saying i should flash both to F10x and my 4 sammys will play together @ 2133 ? (Set both bioses exact same)Quote:
Originally Posted by sin0822
Its only 5.49am lol ive still got time to test. This wud explain the weird behaviour then every time i flash back to F8
well keeping the BIOSes the same is ALWAYs a good idea, but your samsung should be able to run all at 2133, i just don't know if your CPU IMC can do it, but it should be possible with Sandy Bridge, but IDk about 16GB. So if you had Ivy it should be easier.
F10X has a different startup sequence for the memory, thus it allows booting with samsung at higher frequencies, however this startup sequence might not be so good for other types of memory, and thus when this fix is incorporated into a final BIOS it will only use this startup sequence for the generic samsung memory and not the others, however down the line it might be more user select.
That is why i kept saying this is just to test out samsung memory OC.
well keeping the BIOSes the same is ALWAYs a good idea, but your samsung should be able to run all at 2133, i just don't know if your CPU IMC can do it, but it should be possible with Sandy Bridge, but IDk about 16GB. So if you had Ivy it should be easier.
F10X has a different startup sequence for the memory, thus it allows booting with samsung at higher frequencies, however this startup sequence might not be so good for other types of memory, and thus when this fix is incorporated into a final BIOS it will only use this startup sequence for the generic samsung memory and not the others, however down the line it might be more user select.
That is why i kept saying this is just to test out samsung memory OC.
Sin i tried it set both exact same settings in both bioses F10x it wud boot then get into windows but fail shortly after. HCI Memtest detected error in secs. Then when i restart the PC goes crazy into bootloops & beeps. Its funny cos i found i couldnt boot with my normal OC wid F10x got BCCode E1 or soma like that. Back on F8 both bioses and evrything sweet backed down to 4.9GHz for temp sakes.Quote:
Originally Posted by sin0822
I just today got my mates P8Z77-V PREMIUM + 16GB Sammy + i7 3700K He's on holiday for 3 weeks so ive got it to play aroun with. I tried to shoot straight for 2133 (10 10 10 28 1n) it failed in same way as UD5. Its got me thinkin it must be soma to do with 2nd & 3rd timings.
Im glad i didnt get Ivy thing runs toooo hot 4.6GHz needs round 1.375v 85c ++ when i set his to 1866 same timing as ud5 9 9 9 27 1n 1.45v boots fine. Asus bios very refined mouse very smooth but hell its so much to take in too many options make ur head spin.
So, I tried increasing BCLK again...101 MHz no post, 102 MHz no post, 104 MHz running fine and stable:D
There's a posted update for the INTEL Management Engine (ME) controller chip to firmware 8.1.0.1248:
http://www.station-drivers.com/forum...hp?f=32&t=3974
I flash upgraded using the included 64-bit Windows flash program.
Note: must be run from a DOS window command prompt:
"FWUpdLcl64 -f ME8.1.0.1248.bin"
ME firmware should be integrated into the BIOS update.
Yea,8.1.0.1248 is already in F6x (so, should be also in F10x).
The linked .rar file does include an ME version checker.Quote:
Originally Posted by stasio
So: before update flashing, the ME firmware was 8.0.x.x (didn't write it down).
So, this 8.1.0.1248 firmware is definitely newer than what Gigabyte may have attempted to provide via bios update.
There's no mention of any ME firmware updates here:
http://forums.tweaktown.com/gigabyte...fied-bios.html
My Z77-UD5H board shipped with bios version F4, have since flashed to F8, then four F9betas, then F10, and currently F10x.
Note: there's a still newer ME driver version: 8.1.0.1263, but a matching version firmware update doesn't seem to be out yet.
The Intel HECI.inf driver installer from the version 8.1.0.1263 driver package is dated: 2 JUL 2012.
Original F6x BIOS:
http://i3.aijaa.com/b/00348/10720867.jpg
yea please don't try to update ME on your own lol.
I see the GB/HWBot challenge closed with some sick scores... Aristidis 6m55.484 wow :shocked: & Dino, Zen pulled out some nice last minute updates...... 3D challenge also got some big numbers :up:
Sorry: the board referenced is the GA-Z77X-UD5H (running with the latest F10x bios + older ME 8.0.x.x firmwareQuote:
Originally Posted by stasio
[before running the Intel-provided updater] ), not your much newer board, the GA-Z77X-UP5 TH.
Can you provide a screen capture that would actually prove that an ME firmware (of whatever version) is included withinQuote:
Originally Posted by sin0822
any particular board's flashable bios code?
What if: the only method of updating the ME firmware is via separate utility (provided for that purpose by Intel)?
I'm not saying you're wrong, but I'd like to see some proof.
Why is there no ME firmware listed anywhere on this page?:
http://forums.tweaktown.com/gigabyte...fied-bios.html
Did any/all of these bios modders simply FORGET about updating the ME firmware?
:::::::::::::::::::::::::::::::::::::::::::::::::: :::::::::::::::::::::::::::::::::::::::::::
OK, to answer my own question: the Intel ME firmware is indeed contained within the motherboard bios.
With the Gigabyte Z77X-UD5H board's dual bios, I'm running bios version F10x + ME firmware 8.1.0.1248 in the Primary bios,
and F10x + ME firmware 8.0.13.1502 (screen shot below) is in the backup bios.
Attachment 129370
Hi Sin need big favor or you may allready know. I read somewhere that first version of UD5 (Blue cap) doesnt overclock memory so well where as the (Purple cap) version has no issues.Quote:
Originally Posted by sin0822
Is this true? Please find out for us.
Good news: there's a new Gigabyte Z77X-UD5H Bios version F12 out (dated 8/16/2012) that I can confirm: incorporates the latest Intel ME firmware, version 8.1.0.1248.
http://forums.tweaktown.com/showthread.php?t=28441
No notes for F12. May have included some other firmware updates as well, such as the Intel Raid controller, etc.