Printable View
Thanks Andres, I've observed the huge jump in vcore required with each 100MHz bump in clock speed also. Quite disappointing but as long as I can keep it below 1.5 I'm ok with it, especially with temps maxing out around 70 under Prime/IBT. You must have quite the CPU to hit 5GHz with HT and 2800 mem speed! Color me jealous :p
Grr, this is frustrating. I think I've finally found stable settings but I've lost performance in IBT with mem @ 2600 c11 (was able to drop command rate to 1t though). Was at a constant 140GFLOPs in maximum stress test mode before with mem @ c10 and now it's "only" 137. Bclk seems like it's not going to be adjustable either. It's not the worst thing that can happen or anything, it's just disappointing to have spent more on a replacement board and somehow lost performance... Any thoughts on other timings I can adjust to regain that performance without likely affecting stability? I'll test of course, just looking for a place to start testing is all. Thanks.
Imo, try other bench programs like hyper pi or even super pi 32m or even gaming (with oc'd setting) if it will make up a bit of oc dissapointment.Quote:
Originally Posted by techguy2k5
Just my 2c:)
heh, yeah that's an option but now it's not stable anymore either :(Quote:
Originally Posted by Dumo
This is so strange... I keep playing with LLC and fixed voltage and it seems like the voltage fluctuates more depending on what fixed voltage I set. Right now testing @ 1.44 and holding 1.44-1.448 under Ultra High LLC, seems to be stable. Extreme LLC pushes that up by as much as another .04, pushing temps up another 3-4 degrees unnecessarily. I think I'm back on track for stability but the performance drop bothers me. So many variables here, can't figure out why it would've been stable before and not now...
And randomly it's passing again, with voltages no longer fluctuating. Something's not right here. Could it be BIOS? I saw the 0904 BIOS is preferred and I flashed to that at one point but thought it was responsible for the huge voltage spikes in TurboV so I switched back to the latest 12xx BIOS (BIOS not responsible for voltage spikes, BTW).
@tech
what i do its use BF3 to tune the OC...see BF3 will not load maps and will error if not stable, then you should play if you cannot complete a game then u need still a bump, when you are able to game for some hour then you are done...i prefeer this method than prime
and...as my friend Dumo says first steps try hyperpi and other programs then you go to game, prime its for under 5GHz inmo
then when you are ok with LOAD vcore you put it offset and you can also make profiles on AI SUITE for GAME, NORMAL etc...
good luck
Cheers !!!
Sergio
Andres, my only real use for this machine is flight simulator, so extreme stress tests are probably not the best way to test, as you suggest. However, I can't risk crashing in the middle of a long flight or corrupting my 100GB+ FSX install with many add-ons so I have to be completely sure of the stability before I "put it into production", so to speak. I'm going to start over and see what I can do, backing RAM off to default timings and 2600 clock disappointing as it is. Thanks for all the help to both you and Dumo, I'll report back results when I'm all done!
Update: dropped RAM all the way down to 1600 and started testing all over again with CPU @ 4.9, was getting errors in IBT so upped vcore from 1.330 to 1.335 in TurboV and that was all it needed. Going to start testing settings necessary to bump up RAM speed again and hope for the best.
Turns out it wasn't stable after all. With my old board I needed right around 1.48V vcore for 4.9GHz and I was hoping to need a little less with this board due to it's excellent power regulation, I guess that's just not going to be the case as I continued to get infrequent WHEA errors and crashes even at a seemingly stable voltage around 1.45-1.46. So I've started over and am isolating the elements. Focusing on RAM speed now at around 2650 (2600 in BIOS and bclk @ 102 via TurboV) with 11-13-13-28-1t and stock volts, auto vccsa & vccio.
Thanks again for all the help, I think I'll get to my target RAM & CPU speeds but it's just going to take some doing.
inmo you are wasting your time using IBT or LINX and also degrading your chip
just try FLY SIM and thats all
I see people are still waisting their time and precious cycles trying to figure out how to lower the vcore with 0.01 sigh... if that is gonna make a world of difference in power consumption or heat output... and not even one of the best boards around is gonna lower Vcore demand with 0.1, you have me really puzzled techie...
I'm focusing more now on making sure I don't have errors due to running the RAM at settings it can't handle, I'll settle the CPU OC back in around 4.9GHz at whatever volts it needs (hopefully still less than 1.5). I wasn't expecting a full tenth of a volt drop from the MVG, btw, more like a hundredth or two.
I just don't know what to think of this board anymore. I could always run HyperPi at 2666 C10 without trouble. Then one day out of the blue this isn't stable anymore, i even had a hard time getting HyperPi to finish at 2600 C10. Tried all sort of things, tried several bios' but changing over to the red ram slots helped a great deal. Suddenly it ran SuperPi 32M at 2800 C11 LB 6.
Then the next day when i try again it's not stable anymore. wtf? Changing ram back to the black slots didn't seem to help. 2666 seemed stable with the rams in the red slots but when i try to boot at 2800 the board keeps rebooting! Then i try to boot at 2933 and it boots straight into Windows! I could even run SuperPi 1M at that speed. After that i was able to boot at 2800 again but still not 32M stable.
What the f is going on with my board? I'm using bios 1204 atm. Anyone else with similar experience?
http://img.photobucket.com/albums/v6...ps19e4d89a.png
w0w ...think I'm still on 078 or 9 somethin' ........... :D
Have you taken it cold?Quote:
Originally Posted by Zeus
I had one that was a tank, and the other one broke after one session of cold no kidding, one day it would not turn on however there was power to the board and all leds were lit up. Also I was able to clear cmos but the power on button nor the header was able to start it, so just weridness all around sort of like your issue. Andressergio had the same issue I believe, maybe something to do with retraining IMC during booting with different straps / settings?
what vccio and vccsa?
Tried that one either Pete, turned out to be one of the best bios' i've tried. ;)Quote:
Originally Posted by Expat GriZ
l0ud, i think both were set somewhere around 1.15V but i've tried raising them both without luck. Haven't taken the board cold.
Gonna try clear cmos now and start over again.
Swapped my RAM out for 2600 C10 instead since there's no point in having 2800 stuff with my chip. Trying to OC again and now the system restarts during heavy stress testing (IBT Maximum). Any thoughts? What BIOS setting could be responsible for causing the system to restart when hitting heavy load? It's not over-heating, max. temp is ~75-80 @ 4.9-4.95GHz with voltage 1.48-1.51. OCing using Turbo V, which worked fine up until the RAM swap out, have done plenty of RAM testing and am running it basically at stock (BCLK 101 so 1% OC). I turned off all power saving features except C1E, including disabling Speedstep and all other C states.
Re-tested without using Turbo V, forcing only 3.9GHz from BIOS. Still shuts down during IBT maximum. WTF?
LLC? If you don't set it to ultra high it will droop and might go lower than what your cpu needs.
Anybody else who tried Bios 1302 ? Known Issues ? Feel free to join and report.
Download:
http://www.uploadarea.de/upload/upzu...u4qoej9vy.html
Where did you find it?Does it have any release notes?
SB got it from Asus Support due to some serious Post-issues with 1204 - it works i have flashed it - thought i share it and let you guys try it too.
Nice :) What issues does it fix?Would help for others who have simmilar issues.
any changelog and data about 1302 bios?
if i had one i would have provided it ... sorry guys.