Is that F5 final release, worth it to upgrade from F3D?
Printable View
Is that F5 final release, worth it to upgrade from F3D?
HMMMMM it's better than F5e ??Quote:
Originally Posted by stasio
hmmmm
flashed and work very well!
What is the TouchBIOS?
windows based software able you to get control of the bios in windows.
is it working ok?
yes but only with latest f5 bios
Aside from the touchbios, is it worth the hassle to upgrade to F5 coming from the F3D.. any new fixes from F3D that makes F5 worth it ?Quote:
Originally Posted by Andrea deluxe
yes men!Quote:
Originally Posted by Demo
is better in all
What is "all" for you?
Better overclocking/stability over 5.4ghz? Better hypers clocking?
my better is:Quote:
Originally Posted by Bullet92
peripherial compatibility
ram compatibility and xmp profile
daily ultra stable clock @4500mhz
sli working (stability)
voltage stability
IS ENOUGH?
Of course, but I'm looking into good XOC bios which needs "normal" voltages for clocks higher than 5.4GHz (I've 5.65ghz air chip ;) )
you can easily try.......Quote:
Originally Posted by Bullet92
With temp above 35°C outside, I don't think so :D
lolQuote:
Originally Posted by bullet92
Ok I'll try the F5 final release. Also is there a way that I can get to use the 3x USB power with a non Apple device?.. I kinda need the 1Amp charge rate..
you can only use 2 front red labeled usb ports @1AQuote:
Originally Posted by Demo
Yes that I know, does it work even if its a non apple device as the website had a fine print that its only for apple devices.Quote:
Originally Posted by Andrea deluxe
work with all devices... is a simple overcourrent port.Quote:
Originally Posted by Demo
Just got my new SB system setup and running this past weekend!
Lettin' it settle at stock and crunching for a couple days before I start OC'ing.
great.Quote:
Originally Posted by bluestang
you can try directly after disabling turbo and energy saver:
4500mhz with 1.370v with loadline calibration on and cpu internal pll overvoltage disabled.
I reckon that 1.37V is a bit of an overkill for 4.5GHz overclock. Also, AFAIK CPU Internal PLL Overvoltage should be enabled. It may help you reach higher clock speeds at lower Vcore.
Cheers,
@flood
your AFAIK statement, did it help with your oc?
Yh I think it did.
Anyway, read Sin's ultimate SB OC guide. He elaborates a bit on CPU PLL Overvoltage in the guide.
i have 26k at 4801 mhz settings lynx stable with 1.50 vcore ( cpu-z ) 1.48vcore bios ( F5) what is the real vcore ?, that say bios or cpu-z? all with good temps of course and watercooling and -2 C ambient here in buenos aires LOL!
F5 Final............... any bugs ?
DID touch bios show the same numbers on BIOS ?
I'm away on work travel right now, but as soon as I get home I plan on putting F5 final through its paces. I've seen quite a few posts regarding it as the best BIOS for the UD4-B3 going right now so I have nothing to lose but a few minutes to give it a shot.
I'll post some numbers up soon, and in the meantime keep posting results if you are on F5 final and find it to be a better bios than the others.
Thanks
Yh, enabling CPU PLL Overvoltage causes reboot loops. :( I have sent Lsdmeasap a PM regarding the issue but he hasn't replied yet.Quote:
Originally Posted by Gaul
That's odd. I know a couple people who have flashed to this BIOS and did not have this problem. Did you ensure the keep DMI data option was disabled and to do a complete CMOS reset after flashing the BIOS?
Let us know what you hear and I'm hopeful this is an isolated case (no offense).
What... really?! Bloody hell... Keep DMI data was disabled and I even shorted the clear_cmos jumper after flashing the BIOS. Are you sure that those people use CPU PLL Overvoltage?Quote:
Originally Posted by Frost_Bite
Anyway, sth might be wrong here since every time I flash to a new BIOS it seems that I need a bit more Vcore than before to maintain the same OC. (4.5GHz) :confused: I always use QFlash to flash the BIOS. What should I do... reflash?
Cheers,
First test F5
http://img221.imageshack.us/img221/1736/5ghf5.jpg
Still GOT BSOD on Browsing
So do any of you guys rely on just SVID to work properly to do you OC?
Meaning just set CPU volts to Auto, set RAM volts to proper spec, everything else to stock volts, and then just start upping the multi?
bha...Quote:
Originally Posted by Gaul
try 1.420v (set 1.4500v)
is that possible no BSOD on Hyper PI but BSOD on Browsing ?Quote:
Originally Posted by Andrea deluxe
agrrrhhh !!
Of course it is. It's usually like that.
I've just asked one of my mates who has the same board and he said he'd had the exact same problem. At the moment if Internal CPU PLL Overvoltage is enabled it causes a 3-cycle boot loop upon start.Quote:
Originally Posted by Frost_Bite
r u powering the rig up from cold? i meant u have to switch on the PSU first before pressing the power buttonQuote:
Originally Posted by FL00D
No. Why would I do that? :)Quote:
Originally Posted by Fatbird
Man, OCing these 2600k's are alot different than the 45nm Wolfdales I'm use to!
So here is where I'm at so far...F5 BIOS and I'm relying on the SVID and VRD12 system to work like it should. I have all of the following either off or disabled -- XMP, Real Time Ratio Change in OS, Turbo Boost, C1E, C3/C6, EIST, LLC. Internal PLL Voltage is enabled, HT On, DRAM at 1.60v and all other volts an Auto.
With this my voltages set for me by SVID (Auto CPU Vcore) it works out to this (Multi / BIOS / CPU-Z idle / CPU-Z load):
34x / 1.200 / 1.176 / 1.176 / ?
38x / 1.315 / 1.284 / 1.284 / ?
40x / 1.320 / 1.284 / 1.284 / ?
45x / 1.340 / 1.308 / 1.308 /1.248-1.260 (Prime/LinX Stable) highest core temp 58C (25C ambient)
50x / 1.360 / 1.332 / 1.332 / 1.320 (Prime Crashes)
The highest I tried/dared myself to boot at with with manual Vcore was 52x, LLC On with 1.425 BIOS and I got into Windows, idle was 1.416, passed 32M Super PI (6m48.141s) but crashed system instantly in Prime.
Now, is it worth trying to see what my highest boot into Windows multi is just for kicks? I'm comfortable with a 4.5GHz 24/7 cruncher now, but 5.0GHz would be nice!
Also, anyone here running a 5.0GHz 24/7 with C1E, C3/C6, and EIST enabled? With or without LLC?
Thanks!
I flashed my GA-P67A-UD4-B3 with bios F5 (thought I'd give touchbios a try) and noticed something odd with speedstep started happening. Maybe most people disable the Cstate, EIST, and SpeedStep features, but I noticed that after returning from sleep things ran a little slower than my 4.5 GHz 2500k CPU should. Upon further investigation with RealTemp, i7 turbo GT app, and CPU-z I could see that, after the computer went to sleep (presumably S3 as the bios is set to do) that the multiplier became locked at 16x and never jumped up again until I rebooted. I cleared the bios back to failsafe defaults and re-flashed with F5 and had the same thing happen. Maybe I'm just unfortunate, but I thought I'd report it in case anyone else has encountered this problem. My overclock is prime95 stable but sleep may put special stresses on the system for all I know.
I'm headed back to F4 bios to see if that clears things up. After a few runs of the touchbios program it stopped launching (well, the process crashed and seemed to keep 25% of my CPU occupied until I manually killed the process).
just flashed to F6A..... and testing....
Just an update on my 16x multiplier lock after resuming from sleep (in case anyone encounters this).Quote:
Originally Posted by jwfowble
F4 bios didn't change the 16x multiplier nor did adjusting most of the various EIST/C1 and S1/S3 sleep states. I had manually re-entered all my overclock settings so I went back to system defaults and noticed things resumed from sleep just fine. I went back to F5 bios and am headed back up in overclocking with similar bios settings except with Real Time Multiplier Changes in OS set to disabled and things are working fine with sleep resume up to 42x so far. I'll keep playing with things and report back if I ever find what setting seemed to be the trigger.... right now I'm thinking unstable OC (though fine for 2 hr prime95) or real-time multiplier change in OS being enabled. Weird.
Update: And the culpret seems to be that enabling real time multiplier change in OS caused the multiplier to lock to 16 after sleep. It's a bios option I only did after deciding to try the touchbios program, which became buggy and crashed every time I tried to start it after only a few uses. I can't think of any reason to enable this outside of trying a windows-based overclocking program so I won't miss it.
F6A link available please?
http://forums.tweaktown.com/gigabyte...beta-bios.htmlQuote:
Originally Posted by hersounds
thanks again stasio
Hey up mate,Quote:
Originally Posted by Andrea deluxe
Is it any good?
Cheers,
no issues here!Quote:
Originally Posted by FL00D
same clock @ same voltage.
any news over F5, differences in bios or something ???
GA-Z68X-UD4-B3 - F9c
-improve boot loop issue/cold boot start
-01.Sep 11
I just flashed F5 last night, I dunno.. probably nothing has changed.. whats supposed to be fixed with F6A?
EDIT:
My other PC with the same motherboard always gets a bios check sum error usually after from a shutdown or If I add in/change harddrives, updated it already to F5 official and also with the back up bios to F5... this PC is not overlocked and is always running with the "optimized defaults" configuration, its actually just running as sa server ( doing nothing net )
with a 2600K, Corsair 1866Mhz 8GB ram ( only running at 1333Mhz to avoid issues ) 750Watts silverstone PSU, some 9500GT for display..
Is the bios chip or the motherboard it self defective?
changelog?Quote:
Originally Posted by stasio
http://forums.tweaktown.com/gigabyte...-bios-508.htmlQuote:
Originally Posted by Andrea deluxe
Btw,
-Improve boot loop issue/cold boot start
Anyone know why my other board almost always says bios check sum error or overclocking failed after a proper shutdown, its not overclocked and its been running with optimized default settings, currently using official F5.. defective board/bios chips ?
tryed----
the great is: wow multilevel load line calibration and faster boot
the bad is: work like a :banana::banana::banana::banana: (llc)
back to f5
lol, someone is still using this nasty stuff...gigabyte should give you a medal :D
it's mean no vdroop ? or ?Quote:
Originally Posted by Andrea deluxe
give me one.....Quote:
Originally Posted by Mafio
asus is better?
with these Multi Level LLC is impossible to get a right voltage in idle and on load....Quote:
Originally Posted by Gaul
too much positive on idle and too much low on load....
i tryed various combination without succes.....
Yes, we had both this UD4 and the p8p67pro... there's no comparison, the asus is a rock even after 6months of LN2... the ONLY problem of the asus is that :banana::banana::banana::banana:ing bios that doesn't work for 3dm01, and they're not going to adjust it.Quote:
Originally Posted by Andrea deluxe
Ah, of course me and mafio, we are speaking about LN2 behaviour, in daily they're all the same or whatever ;)
i use my pc for gaming.... 4500mhz daily.... no issues here....Quote:
Originally Posted by |ron
Ok, in daily no issues, every mobo works fine.
When you do ln2 sessions, you can't boot at a bclk higher than 101, so if you need to go to 106, you need to enter the bios, set 101, f10, enter the bios, set 101,5, f10, etc...etc... until you reach the correct bclk. If the system crashes and you have to reset, you'll need to repeat the procedure, really frustrating and time consuming.
After a few hours@ln2, usually the mobo gives up and you'll need to wait till the next day... no condensation or water, just the mobo going crazy.
There are other issues that I don't remember, but we tried 2 of these ud4s, and except for 3dm01 performances, they really aren't suitable for extreme oc :(
lol, i myself tried 3 mobos.
the first one was DOA, the one i had back from rma (after waited 2 months or more) was "OK" despite having crappy bioses, doesn't boot at BCLK different to 99.8 and don't know how many other problems.
in the meanwhile i threw away another 220 € with the UD5...same issues, same :banana::banana::banana::banana:ty bioses but higher price.
way to go giga, way to go.
du...for du every mobo is ok, 100 x 45, 1.35 volt gg no re...every mobo around there can do it.
for benching bah, dunno, 6 months or more and gigabyte haven't yet addressed the issue with BCLK, their crappy boards are the only ones that need 8 reboot to start at 104 MHz of BCLK cause you need to go 0.5 MHz to 0.5 MHz or you get OC fail...
100.5 --> F10 --> 101 --> F10 --> 101.5 --> F10 --> 102 --> F10 --> 102.5 --> F10 --> 103 --> F10 --> 103.5 --> F10 --> 104 --> F10
shutdown the system...never do this...now you need to redo +0.5 --> F10 for another 8 times.
what a joke. :rofl:
Say an old overclocker " all hardware is the one may be not that's one but sure the next stepping is your work trying and discovering " ....like buddha words but with human soul
Ok, so we're dumb, we can't make a mobo work properly and this UD4 is the best mobo ever... ok...Quote:
Originally Posted by hersounds
It's worth nothing if we say we tried more and more mobos p67/z68 and this one is probably the most bugged, most instable etc... we didn't work well, got it. :shakes:
no man. is just a mobo, nothing more, one can be better that other, 2 mobos, 5 mobos, 10 mobos, 27 mobos can be " bad " ....but 1 only can be surprise to all overcloking scene
allright, but when you must choose were to put your money, I love when someone gives advices based on his experience, to let me avoid a "bad" mobo etc...
This is truly not a bad mobo at all, but has a lot of problems for extreme oc, indeed ;)
You gotta control LN2 temp. From all 2600Ks I tested a few is totally freez bug won't go 48X with 0C evap/base pot temp.
Don't waste your precious time fighting this mobo when using LN2 instead use a decent single stage and you'll be ok.....and try to boot with high bclk is too much headache sometimes...boot up @ bclk 100 and use easy tune 6 to go as high as your cpu/ram can go.
Its a basic sb overclocking for almost all P67/Z68 mobo exept the one that easily burnt its pwm when pushed up to 1.65+Vcore for extended benching
I think sweet spot for 2600K is -45C evaporator or base pot temp
nah, here the problem isn't temperature or something else...it's just this mobo is a piece of crap.
already solved my problems, threw away another 180 € and bought an EVGA.
Quote:
Originally Posted by Mafio
try a cheap z68 gene-z
it kicked the crap of my ud7
people with LN2=3 guy
people with air/liquid= the rest of the world
i do not haved a single issues here on air cooling 4500mhz daily-------
A post code reader card will comes handy......help you to understand why your UD4 get mad when you try to boot up with BCLK 105:)
Well, we have a post code reader, but I can't remember what the code was, it passed some time now :)Quote:
Originally Posted by Dumo
Anyway, don't think every SB has it's "cold limit" (not cold bug) at -45... I saw a few of them scaling until -80°c, even gaining a multi.
It's important to try everytime... then, if it doesn't matter to be at -45 or -80, you're right, a SS is way more "confortable" :)
@zasmx: yeah, it's probably our next step... change our p8p67pro with a gene-z, at least we'll have a complete platform, even for 3dm01, beloved 3dm01 :)
@andrea: yeah, you're right, but hey, Xtreme Systems means we can take in account even the xtreme oc, evaluating a piace of hw ;)
Yes, I'm talking (with at least) 5750+Mhz chip:)Quote:
Originally Posted by |ron
Less than 5650Mhz (wall) chip is just wasting time and LN2 especially for '01 unless peep just want to bench 3D'11
As you know, most if not all 2600K will stop at its Mhz. wall regardless of (subzero) cooling
You still like Gigabyte (thread). :clap:Quote:
Originally Posted by zsamz_
But,please don't repeat yourself everytime same.
Imo misconception with "why I can't boot up with BCLK 104?"...Yes, most of the time (if you use) post code reader will stuck @ C1 or 4E and your UD4 will reboot endlessly.... you wasted that expensive LN2
How bout try BCLK 100 and use easy tune 6 in windows (the mobo maker invented ET6 for a reason)...If you use a good elpida hyper based ram and your 2600K can boot up @ 59X then set BCLK 100/2133/59X in bios then use ET6 frequency slider to go up till max benchable setting your system can go.
For now, overclocking SB basically is to find a 2600K/2500K than can go 59X and mobo should be the last piece of hardware that we have to worry about
If I well remember, mafio tried 3 kits of hypers on this mobo, with no luck: on the asus mobo, it never arrived to post.
On this giga, they arrived to post, sometimes, but no more else.
What kit of hypers did you use, with which bios, dumo?
Its Dom GT 2000 C8, those old TR3 Iron, but it needs 1.8+Vdimm/1.2+VTT to run 7-8-7-24 benchable @ 2133+ with F5 bios. Its too much headache:(Quote:
Originally Posted by |ron
I like MIVE because its smoother on memory clocking and boot up, but it killed my first 58X chip:( Never trust its pwm for high Vcore on extended benching usage afterward
What board do you run now, Iron?
Lol, I just bought used Dominators GT, the ones with psc, not hypers, you're message is not encouraging :D
Anyway, right now I have a dead cpu, so I must rma it and, togheter with that, I was thinking about changing our old p8p67pro B2 in a Gene-z, by adding some money to pay the price difference.
The rog series have indeed soooooo much support, let's hope the gene-z would be as solid as the little pro. :)
P67A-UD4-B3 - F6e is out.
tryed now....Quote:
Originally Posted by stasio
work very well
improvement:
LLC upper voltage in idle is lower than other bios (great)
memory copy up to 1000mb/s greater
aida64 photo worx from 67000 to 69500 (wow)
Z68XP-UD4 - F5a is out.
Edit:
and P67A-UD4-B3 - F6f
This should be new:
http://i10.aijaa.com/b/00384/8694915.jpg
flashed and have the same performance of f6eQuote:
Originally Posted by stasio
but how change the performance that newer setting?
Maybe CTRL+F1 ?:shrug:
or this is for Z68........?
Edit:
try to rewrite the DMI data and Load Optimize Default after flash BIOS.
Btw,
I think in F6f BIOS,GB Taiwan put back LLC....Auto/Disabled/Enabled,so you can use DVID and LLC Enabled again.
whats with LLC back to auto/enable/disable, is f6e not good then or is it just to be there to be found if needed
Read properly above,F6f BIOS have now back LLC with 3 levels (Auto/Disabled/Enabled),as before.
thats what i sad -.-'
GA-P67A-UD4-B3 - F6f
-LLC work with DVID
-19.Sep 11
Btw,
here is same but modded F6f with:
1. IntelŪ RAID ROM 11.0.0.1124
2. Realtek LAN ROM 2.41
fantastic.Quote:
Originally Posted by stasio
now llc work fine..
i have set llc level 4 and set 1.330v and i get:
1.344 idle
1.356 load
Fun!
thx stasio
performace of memory write ,read,copy is great, and the bclk is finally not droping to 99,xmhz its now 100,3 :D
didnt tryed levels yet, i am going to solder wires to condenser and read ii wtih unimetar
Yeah,seems all beta BIOS's after September 14 are getting good performance.
great f**ng bios! ( f6f)
I agree. F6F is allowing me to run 5.0GHz with 1.41v (1.42 loaded) which was impossible before now. I think they're starting to get the BIOS straight and turning this board into a great overclocker.
Thanks for the modded BIOS stasio :up:Quote:
Originally Posted by stasio
it's mean NO-VDROP ?Quote:
Originally Posted by Andrea deluxe
yepQuote:
Originally Posted by Gaul
I really like the F6F BIOS, but I still can't get the board to overclock without using the DVID method for adjusting vcore. If I set the LLC level to 4, the vcore to 1.380 it gives me around 1.404 idle and 1.42v loaded...which is exactly what I need for 5GHz. The problem is that these settings don't pass 10 seconds (literally) of LinX. If I set the vcore to normal and use a DVID offset of +0.055 I get the exact same voltages as above, but I'm able to run 20+ passes with LinX.
has anyone else seen this? I've tried with all C-States disabled, enabled, and auto, PLL override enabled and disabled, Spread Spectrum Enabled and Disabled...but nothing works.
Thanks
Have you compared the temps with both settings? If the temps are lower, but voltage readout is the same (without the offsets) then looks like something is wrong...
Even more important, did you test the voltages with a multimeter? If you test with DMM, then you would know if the board is reporting the wrong voltages or not...and this is important to find out.
Because the old level 4 had massive vdroop, but this one has vrise, but according to your test with LinX, maybe something is really wrong with the voltages, and the vdrop is the same as it used to be back when people had bad vdrop with the lower levels?
i am saying if they follow teh Sniper 2 BIOS as a baseline they will do just fine.
@ all = coz i'm ex UD4 user
i'll buy ASUS p8p67 pro, here's different on BIOS
DIV MEM up to 2400
http://img683.imageshack.us/img683/1447/27612576.jpg
NO VDROOP
http://img193.imageshack.us/img193/1958/extj.jpg