For day to day use, gaming etc is this boaard fine?
Just looking for a mild oc 4-4.2, will this mb be ok?
Printable View
For day to day use, gaming etc is this boaard fine?
Just looking for a mild oc 4-4.2, will this mb be ok?
do yourself a big favor and get a better and cheaper boardQuote:
Originally Posted by n3wbi3
for a 4.2 overclock even a rampage II gene is good enough
i'd consider a cheap asrock extreme for 4.2;)
This is how the boards I have behave for VTT with a given memory kit - sub-timings matched up close as possible. Voltage measured under socket on each board.
http://images.anandtech.com/doci/3820/voltagegraph.png
MSI know about this - I alerted them almost two months ago.
Base layout for this board is a 6 layer PCB, which I suspect uses the same memory layout that was used by MSI on the previous gen X58 boards.
whats program is everyone using to overclock these ??? MSI software crashes on my OS for some weird reason and i cant find the the pll in the setfsb list ?? any ideas ??
Dashboard OC panel
Arrrgghh yeah thats what i been using but it annoys me i wish i could drop bclock or raise bclock then push apply, not as i click it drops or raises :(
If i could use setfsb i could use my macro hot keys to drop and raise bclock mid bench :)
i think the board is underratted. it's working very well since i got it couple of days ago. and for those who bother to play with the xpower, here's the latest beta BIOS 142
http://www.sendspace.com/file/0ms4dv
I bought it too... and i am loving it.
Very good board.
Can yours boot from bios over 220Bclock ? and run 2000Mhz/4000 uncore 32m stable ? Haven't tried the latest biosses... thx for any info
SuperPI 32M with 2000MHz/4000 UCLK is very easy but cant pass LinX.Quote:
Originally Posted by Leeghoofd
Here's my setup with BIOS 1.3
Attachment 107778
no linx pass?
not fun!
Yep, that's the major fault of this board, it's not 100% stable where other boards with the same CPU and memory are. Multi-thread Hyper Pi and Linx are waaay more difficult to pass on this board, needing higher actual voltage than other boards out there.
since ure on this thread raju
wanted to ask u or any other who have done it
hmm on p55 running cpu pll at 1.95 seems to be no issue so far on my ftw for months
but hmm on x58.. is it safe even when i have a good cooling with temps never more than 80c on linx
and whats the worst could happen
ive never hard a intel proc that died from high pll so far.
Not sure how safe it is long term to be honest, depends how much over current the associated parts can take long term. On 775 I know of a few CPUs that deteriorated in absolute FSB ability over time as a result of running about .3v volts over spec.
Bigger question here is though, why do you need to ramp PLL on the Nehalem architecture? Intel include a PI filter on die (among other possible improvements), so hiked voltages don't really do anything for the better in my experience (though there are sometimes gains to be had with lower voltages at sub-zero temps).
-Raja
hmm just checking safety limits while i try to get 2180 2:10 on a 920.Quote:
Originally Posted by raju
whats a PI filter?? care to explain.
I guess you can try fiddling with the voltage and see what it does for you, though I tend to veer away from overclocks that require a voltage hike to this rail.
The PI filter, usually consisting of a couple of capacitors and an inductor, helps filter out hf noise which clock circuits are especially sensitive to.
Ok so I got this board despite everything bad Ive read about it. Just wanted to try it, liked the layout. So far 220bclk, no linx or blend just wprime 1024m because this chip isnt much of a clocker. Ill keep trying but I have a feeling this will be about it for bclk.
If anyone still uses or remembers anything about the board, I had a question. The sensor in hwmonitor~TMPIN1 reads very high, I looked around the web trying to find info but all I could find were screens in a couple reviews and in other forums. Is this sensor borked or is my board really running this hot? shows 65-67 idle :shrug:
Could be the temp of the IOH (put a fan over it and see if it drops)
I just stuck some more fans on the board, temps a little better.
I should have used a chip I was familiar with to try this board, seems my bad luck hasnt ended, both 950's I got are poor clockers, or I just suck with anything other than 920-930's.
Ill try and get this stable tonight, I know the chip is good for it.
The board did 220x21 better than my UD7 as far a vcore and vtt, also better that the e770. 3x1gb kit hyperx.
So far 225bclk, boot from bios. vcore is a bit in high in cpuz but LLC is disabled, QPI 1.417 real tested with DMM, IOH 1.30v, DRam 1.66, GTX460 @ 875/1750/2050
wow OC Nub, nice OC:D
legend said the one would come
and unlock this MB's secrets
i think maybe your the one:up:
Thanks radaja, I am starting to like the board. It was hard to get the bclk up but its not doing bad at all now.
That's a nice one OC_Nub
Some questions :
could you plz give us the revision nr of the board
and some bios screenshots of the voltges and co...
We have 4 Xpowers at the lab and none booted straight from bios over 220. (yes with the soft boot trick, but then you shutdown completely or crash and then it won't post again)
Thanks. I will get the revision number after work today. I saved a profile for 225, but it wont boot when I load it. Only way to repeat this is to start at 221bclk, boot to windows, restart and set 222 and boot to windows, and repeat this process going up 1 at a time until I get to 225, otherwise I get A1 A2 or A5 code (cant remember) and no boot. Once at 225 I cant reboot the pc and it will go to windows fine or even turn it off and boot, but if I load defaults then have to do this process again. I had to do this on the rex3 to get from 115 to 222, but lost 3d performance on rex3 and would not reboot, this board doesnt lose performance. The setting Low Droop for QPI doesnt work great, with 1.50v set in bios it still droops to 1.417v with DMM, at lower voltages like 1.4-1.41 the droop is much less, gives 1.34-1.355 real.
I kept trying last night but that was the only way it worked. I will try my 920 this week and see how that chip does because it is mych better for high bclk. Tried with the 950's and they wouldnt boot at 225, only this 930 so far.
Edit: I goit home and was able to boot a higher bclk without having to go through that process, but its not consistant. Will keep working on it tonight.
The problem is when I change the ram timings, save and exit, like on my giga board or I guess most boards it powers off when you save and exit. Thats when you get the A1 code and it refuses to boot. As long as I dont change timings it will boot at 226 np. So if I want to try for more bclk I just need to set the timings Im going to use before I start testing, then even if it fails and I have to change voltages it will not get stuck at A1, and I can continue to try for higher. if I change timings and try and reboot to 226, I get A1 and have to start over at 221-222, save and exit, go back to bios and set 226 and its fine again, then I can reboot or power off and come back later and it boots fine at the same bclk. Hope that makes sense.
See if i can help out with this board some.
This was extremely difficult in triple channel and should not be.
http://chew.ln2cooling.com/qdig-file..._X%20power.jpg
Dualchannel the board does perform and clock easily however.
http://chew.ln2cooling.com/qdig-file...rg_Xpower2.jpg
Board will be on the bench 24/7 for the next month, hopefully we can get it ironed out.
Nice scores chew.
I tried some more and it consistently boots up to 233bclk (high as I have tried)as long as I dont change timings. Once I do make any changes to timings I have to set 221 bclk or less with timings I want to use, boot, then I can go straight to 233 so far. It will reboot fine, I can change bclk up or down change anything but timings and no issues.
None of this is stable, and not gonna break any world record with these settings, but just trying to get the hang of this board. These were boot from bios, no software. Voltages were just a guess, could possibly be lower.
oc nub that is downright sick:shocked::shocked::shocked:
what the f :shocked:Quote:
Originally Posted by OC Nub
Thanks.
233 is as high as Ive been able to boot. Once I get the screen saying "Previous Overclock has failed", it reverts timings back to 8 8 8 19 trfc 59, and have to start the oc over. Still not bad, the board isnt strong with memory like the e770 but I havent spent much time on it, maybe with more practice it will get better.
Trying to get the timings down and rase the uncore frequency. Its very hard to boot setting the uncore ratio to 10, at least it was for me. I had to enable to over voltage jumper for qpi to do this, most it will allow me to set is 1.60v bios, which only gives about 1.50v real. I stopped after this, will work on it again later. I dont have the cooling to do some of the benches I would like to try, but Im pretty happy with the results so far.
@ Leeghoofd, where do I find the rev# on the board? the numbers I see are 01BA1608 and A916050547.
With a little tweaking the board is doing very nice. I took the Classified out of my main rig and installed this XPower, it is more of a challenge than the Classified, but with some work it will get the same results with less vcore. Also the 3x1gb kit is faster to test linx, tired of the 1hr+ linx runs.
@Leeghoofd, I read the review of high end x58 board at Mad Shrimps, nice review. I thought it was a very unbiased and fair review. :up::up:
I was able to get the temps down on the board from 67 to high 30's low 40's by taking the hs off, raplacing the Paste, also putting the board in a different case helped.
awesome :up:Quote:
Originally Posted by OC Nub
Thanks Sergio.
Got a new kit of Mushkin Redline 998790 1600 6 8 6 24 to use as a backup kit. They clock pretty good, better than my hypers on this board.
Got a second kit of Mushkin ram, this time I tried the Radioactive 1600 c6. This board really seems to like them. I could not get my Dominator GT's to pass 32m even at 2000 8 8 8 24 without jacking up the voltage. The E770 did great with the Hypers but not good with the Mushkin.
is the qpi droop that bad?
same thing with my rexII it does much better with bbse n powerchips than hypers
It has a lot of droop even with Low Droop set in bios, 1.49v QPI gives like 1.41 1.4125 load.
man thats freekin ridiculous for a newer boardQuote:
Originally Posted by OC Nub
n i thought the rexII droop was bad lol
I have 1.55vQPI set in bios now, idle is 1.485, and load is between 1.462-1.467v. Not for this run though, QPI was much higher.
Pushing the ram a little more, 2000mhz 6 9 6 is about as far as it will go. This isnt the best chip, vcore is high but it runs cool.
Playing around with uncore, this is about as high as I can get it to go and still complete 32m. Forgot to lower vcore for 4.4ghz oops.
http://img262.imageshack.us/img262/9392/uncore4600.jpg
Uploaded with ImageShack.us
that is a crazy nb speed:shocked: wish i could run my ram like that....Quote:
Originally Posted by OC Nub
but maybe one day i will:shakes:
OC Nub, how old is your board? Did you just purchase it?
This might be a bad thread res, but have these major bugs been fixed or at least improved? I'm getting the board and will be putting it in a liquid cooling system... which I haven't really heard to much about with this board. It seems that it's not made for air, but extreme cooling isn't doing so hot either, maybe the middleground might be better? I'll probably be keeping the i7 930 at 4.2GHz because of the apparent stability issues.
Thanks,
ScottALot (a lot of Scott)
Ive had the board about a month or so, dont think its a newer rev.
As far as being limited in any 24/7 oc I dont see any issues withthe board at all. It will clock as high as my e770 or UD7 24/7 and requires less voltage. If you are limited to 4.2ghz on your 930 it isnt because this board is not capable. I havent found any stability issues with it for 24/7 use.
I dont run raid, or use the sata6gb/s, or the USB3.0, or use a lot of other features people may want in a board. But for the things I do it has been very nice.
I think as of right now its my favorite x58 board of the ones Ive tried, it recovers from a failed oc better than other boards, and requires less voltage for my 24/7 oc.
you must know how lucky you got;)
one hell of an IOH on your board:up:
Quote:
Originally Posted by OC Nub
^Thanks, very happy with this board.
One hell of a lot of QPI voltage to :rofl:
wanna lauphQuote:
Originally Posted by OC Nub
i'm playin with an msi x58m
got her to 230blk:rofl:
not bad for a micro board lol
my xpower couldnt catch her:rofl::rofl:
Sorry I don't have the board anymore (wasted mine with the latest beta bios a while back) Will ask Massman to take some pictures, there are added mosfets and co on the newer rev. But not sticking my hand into the fire if I tell something wrong...Quote:
Originally Posted by OC Nub
Will come back with info...
Thanks man, was a bit dissapointed when the engineers of Asus at the tech conference, said that Ocing is not warranted... The issues you and Zsams had should have been adressed via a mobo swap... Glad you like the Xpower; yours is running spot on :)Quote:
Originally Posted by OC Nub
Nice bclk, I need to try a micro board. Hard to imagine something that small clock so well.Quote:
Originally Posted by zsamz_
Well I haven't actually used the i7 930 yet, so I don't know its clock limits... I just see 4.2GHz a lot and it seems like the best ratio between overclock amount and issues present.Quote:
Originally Posted by OC Nub
Quote:
Originally Posted by OC Nub
its just my luck it has a great IOH
but they need to fix throtling issue:mad:
The revision number should be just below the 2nd PCIe slot. Should say something like "MS-7666 VER: 1.1".Quote:
where do I find the rev# on the board? the numbers I see are 01BA1608 and A916050547
@OC_Nub: Great results there...
Thanks. "MSI-7666 VER:1.1" is what this board says also.Quote:
Originally Posted by maxxx
Whipped out a mod or 2 on this board related to VTT.
Now set 1.30 bios = measured 1.325 DOS based memtest.
Gonna fool around in memtest a tad more than installl 64 bit win 7 and see what she can do.
Edit.
Ok fine tuned mod, gonna chat with MSI see if I can "release" it.
Basically once fine tuned 1.32 set in bios = 1.32 load in windows all cores enabled running 32m.
Ive never tried to mod anything before. Looking forward to giving it a shot if you are able to release it.
I got permission.Quote:
Originally Posted by OC Nub
Mind taking a high res shot above the 6 pin power adapter?
I need a shot in between the 6 pin power and the little mosfet sink.
Mine is all hot glued now so won't really show what i need.
Here's an example.
http://chew.ln2cooling.com/XS%20mods/vtt%20mod.jpg
http://chew.ln2cooling.com/XS%20mods/vtt%20mod%202.jpg
with mod, VR set to 40 ohm 1.35 set bios measures 1.35 real.
http://chew.ln2cooling.com/XS%20mods/vtt%20mod%203.jpg
damn thats a tight spot. did you remove that small resistor?
yes.
I've solder to smaller and in tighter......this is rather simple.
Preety satisfied with the way the mod is working out.
Should help 24/7 users greatly.
Little dissapointed in what i am seeing with uncore performance however.........could be due to the fact that it's bottlenecked by the crappy ram i'm using atm.
http://chew.ln2cooling.com/qdig-file...%20mod%206.jpg
http://chew.ln2cooling.com/X58%20Xpo...%20mod%204.jpg
Had to chop something out of this shot. Will add it "unchopped" at a later date ;).
As you can see the gains from a rather high uncore clock is dissapointing.
http://chew.ln2cooling.com/qdig-file...%20mod%205.jpg
Thanks for posting the mod. I think this is above my skill level though.
Nice Uncore!
is there an official bios that supports the 990x?
msi.com doesn't have it yet
Maybe you can find one here : http://forum-en.msi.com/index.php?topic=140522.80
@maxxx are you using 990x already ? :D
@bart: nope, not yet...Quote:
@maxxx are you using 990x already ?
thanks i'll check it out tonight..
I have this board, got it with a i7 950, can do 4.2Ghz with 1.28V and the only problem is that my 6GB OCZ P 2000Mhz requires about 1.54/1.56V QIP on this board.
On the resent bios update V1.4, you can only set qip v up as far as 1.52V(qpi) which isnt enough for my Ram, for now the system i have it on standby. i may just sell it, dont want to buy more ram.
besides that, great board.