Printable View
Quick question. Do we know what Gigabyte does differently (any clues from their bios?) to achieve higher max boosts at lower settings with their GTX 660ti OC model (GV-N66TOC-2GD)? Seems to be the case in this and other reviews:
http://www.xbitlabs.com/articles/gra...i_3.html#sect0
p.s.For those who cant extract their bios with GPU-z, heres a database of bios @techpowerup.
http://www.techpowerup.com/vgabios/
I did not really see what the boost was. However what I can see is the memory clock on this card is through the roof at 7468mhz, this probably accounts for most of the enhanced performance.Quote:
Originally Posted by aamar
I'll take a look at the bios.
Its stock boost (1111mhz) peaks to 1293mhz under load while the Zotac (same boost clock, 1111mhz) only peaks to 1215mhz. The Gigabyte card has a higher peak boost than all other 660ti's (about 182mhz vs around 100mhz for other 660ti's). This 'extra boost' was also seen in other reviews I believe.Quote:
Originally Posted by CrazyNutz
Ok then this is because of the max boost value in the bios. On the gigabyte it is higher than others letting it boost higher. I havent checked the bios for this card yet, i will soon to confirm.Quote:
Originally Posted by aamar
I do know where maxboost is in the bios, and KGB will support this value very soon.
Awesome! :DQuote:
Originally Posted by CrazyNutz
aamar+1 Cant wait, I may actually put this card under water now, Seems worth it.
Ill just leave this right here.
http://www.overclock.net/t/1341231/kgb-5-virus
I dunno whats up with the task scheduler, I assure you I do not do that nor do I make any windows API call's directly at all. KGB was written on linux, and ported over to windows with MinGW.
The virus/trojan Is a false positive. I compile with MinGW, and software compiled with MinGW has been known to get false positives from some virus scanners:
http://comments.gmane.org/gmane.comp...ngw.user/38529
http://lists.freedesktop.org/archive...er/024866.html
http://sourceforge.net/tracker/?func...35&atid=202435
http://www.mail-archive.com/caml-lis.../msg02842.html
fyi,
kgb.exe 0.5
sha1: 263170338ade617c37093e80684559b6e2c7e24a
md5: f133d2e6938d062fb879f462a4cb8eb8
the binary is clean (checked with Avira). virustotal.com is a useful site but only reported heuristic type of alert which should be taken with a grain of salt.
CrazyNutz
I simply wanted to give you the opportunity to resolve the issue on your own, which you have done well, its the only reason i posted a link. The last thing we want on our hands, is an epidemic across the internet.
Keep up the good work.
Yes thanks for bringing that to my attention :) That certainly could have gotten out of hand.Quote:
Originally Posted by KGB7
CrazyNutz
I also received a message in my Inbox, which is intended for you.
Quote:
Originally Posted by krisvonsiksteen
Thanks. It looks like he needs a tutorial could you reply back to him with this: http://www.xtremesystems.org/forums/...=1#post5159371Quote:
Originally Posted by KGB7
His/Hers, mailbox is full.Quote:
Originally Posted by CrazyNutz
Guys some manufactures are releasing UEFI vbios for keppler cards. This is for windows 8 fast boot.
It has been brought to my attention by users error-id10t & V3DT at Overclock.net that KGB is not working correctly on these bios.
I have created a new version https://www.dropbox.com/s/fqy9txhgca250qi/kgb_0.5.5.zip
I need some testers with a card that has a UEFI vbios.
Also this version will show your max boost clock. (can't change it yet, but that's just around the corner)
Any luck with that boost cap value? I think that's critical in making any OC software redundant.
edit - nvm just read the above post!
CrazyNutz, thanks for all your work on this! I gave 0.5.5 a try thinking that my Asus GTX680-DC2-4GD5 would take it. While I get no errors in the process, I get a Code 43 whenever I restart with the modified BIOS. I've done a bit of searching and I'm pretty sure I'm the only one to have this happen. :(
Throwing you a copy of my BIOS, in case it's a problem anybody runs into in the future: http://db.tt/y6qYSdiC
In addition to the max boost clock setting it would be great if we could get a base clock setting and memory speed setting. I personally like to test the card(s) and then flash the bios with stable clocks and not worry about software other than monitoring :)
Thanks for the effort put into this already it's greatly appreciated :up:
keep the good work crazynutz and happy new year !!!
and for all !!!
Yes, the code is in, I'm just trying to get the uefi thing out of the way first.Quote:
Originally Posted by CryptiK
crazynuts
the KGB doesnt touch the original core and mem clock right ? can i edit the .cfg to get till 1,24volts ?
I think I figured out the problem with your bios. Try this one please: https://www.dropbox.com/s/4upapvqcrdujr7c/kgb_0.5.6.zipQuote:
Originally Posted by Avalonstar
Just curious will the voltage adjustment eventually be able to stick or am I just doing it wrong. I've changed the values but still only get to 1.21 (1.24 hardware). Which btw the difference between software and hardware is a pretty much straight number. It's about .04-.055 which is just the variable range from the power draw and droops. I have tested this thoroughly because I have heard too many times that afterburner doesn't report the correct voltage. Afterburner indeed does report the correct software voltage. The hardware voltage is like I said usually an increase of that increment I mentioned.
1.2125v software is max doesnt matter what you put in the config file higher.
For reference cards yea I believe that just wondering how it affected Lightnings and other cards that were built to have voltage control. I used to take my old Galaxy card to 1.3 no problem just wonder if this mod will allow the same.Quote:
Originally Posted by CryptiK