Well, replace the NB heat sink with a better one, like n old alpha Socket A cooler with a decent fan. Im running Vmch @ 1.95V (real) and uguru never reports over 40 degrees.
Printable View
Well, replace the NB heat sink with a better one, like n old alpha Socket A cooler with a decent fan. Im running Vmch @ 1.95V (real) and uguru never reports over 40 degrees.
I'm well happy that i've now got to 390x9 thanks to your advice. Voltages are a lot higher than i wanted but i guess its whats needed.
vMCH = 1.8
vCORE = 1.55
vDIMM = 2.45
Tightened the timings as well, all in all i am pretty happy, the only thing im concerned about now is the temp, uGuru is reporting the cpu at 50 and i guess thats at idle, i need to gauge it at load, i dont think my scyth ninja is doing a very good job.
Got a couple more questions for you guys and gals
I am using vista, does anyone know of a decent core temp alternative?
Speedfan is saying it can not find a supported device, whats that all about? i have an e6600!!i have disabled all the advanced CPU settings, would this cause a problem?
Many thanks to all :)
Steelsie
Do not trust these onboard sensors, my new L628B Xeon sees 83+C in coretemp priming :( 3903mhz/1.58v realQuote:
Originally Posted by Steelsie
Old L629 Xeon saw ~66c priming 3903/1.63v
Mobo walls for me priming @ 435fsb @ 1.91v real (2.0 bios) vMCH and a MCW-30 on the chipset...
thanks for your advice :)
testing now with more vcore.. yes the NB cooling is the next thing i would do.
My build is finally complete!
http://static.flickr.com/105/308179968_ce8497a798_b.jpg
http://static.flickr.com/102/308171600_fe5fbdfd8a_b.jpg
Final with cable managment.
http://static.flickr.com/110/308176514_9d02916024_b.jpg
You can check you all the pictures here
http://www.flickr.com/photos/amx/set...7594395210360/
anyone know how to change pci-e w/o usuing uguru?Quote:
Originally Posted by Ten
everytime i try, my computer locks up and then i can't boot up again until i clear cmos...
in uguru setup/OC guru and u can make the change, i'm getting the same but that's because i forgotten to change from 4/5 to 1/1 in bios memsetting when i'm runnig in 401fsb:) bwt with old bios i think 10.01 i don't remember i had the same problem but after flashing no problemos other then i forgot memsetting:)
PCIe frequency adjustment has never worked on this board, I was hoping the new BIOS might have addressed this.Quote:
Originally Posted by shimmishim
PCI-E can be adjusted easy with a PATA HD.....Not a SATA one.....Up to 142MHz for me.....For a SATA HD to handle up to 120MHz, you need the VICH-VTT mod....;)Quote:
Originally Posted by G H Z
so where is the option to do so?Quote:
Originally Posted by hipro5
are you saying that pci-e adjustment is only allowed if i have a PATA hard drive installed and pci-e adjustment will disappear with a SATA drive installed?
to amx
m8, your mobo is in btx position.. you should know that the silent otes are not made to work on btx setup..
so i think you need to replace your chipset coolers to custom ones.. what you have there gets very hot on the mobo when oces ,in the btx position
Νο......With the latest bioses OR the latest uGuru, there in NO PCI-E option......You must use the previus version of the uGuru so as to see and change the PCI-E frequency......:)Quote:
Originally Posted by shimmishim
okay. so you've answered my question.Quote:
Originally Posted by hipro5
there is no way through the BIOS to change the pci-e frequency.
i guess i don't have the latest uguru since i can see the pci-e frequency and change it.
however whenever i change it, my computer blue screens and windows won't boot. i have to do a cmos clear in order to get it to boot.
There's is an option in an older official bios (first release if I remember right) that has the option into bios....:)
IF you install a PATA HD, you'll be OK and you could change it....but with NO SATA on the mobo.....:)
http://www.thelab.gr/images/Hipro5/A...26MHz_bios.jpg
http://www.thelab.gr/images/Hipro5/A...32MHz_bios.jpg
TIP: DON'T clear c-mos every time your mobo crashes......POWER OFF the switch behind your PSU and after a few seconds (you'll see mobo's stand-by led goes off) , switch it on again....and power on.....;)
.
ahh gotcha. that makes sense and thanks for the crash tip.
do you think if i plugged in a pata drive, set it to 101 and then installed a sata drive, the pci-e frequency would stick?
No SATA.....at least for 3D benches which we need at least 125MHz PCI-E for the VGAs to unleach their power...... ;)Quote:
Originally Posted by shimmishim
...and no SATA for higher PCI-E frequencies....:)
Beautiful rig, nice work :toast:Quote:
Originally Posted by amx
My board crashes on BIOS 1.02 with 1mhz frequency change (SATA), the shipping BIOS was good for maybe 10mhz. I'll try an IDE drive and see what happens, thnx hipro ;)Quote:
Originally Posted by hipro5
WOW. I think I FINALLY FOUND THE ISSUE TO ORTHOS BLEND TEST on my system....
I think I may have figured out part of my problem.
The Crucial Ballistix PC6400 are rated 4-4-4-12 @ 2.2
I run 9 x 401 with 1.45 vcore + 2.2 vdimm.. SUCCESS!!! I got orthos blend to run for 20 minutes, something I've never done before.
So I get a little greedy, up my vcore to 1.5 + 2.3 vdimm and 9 x 413... blend errors at the usual 4 minute 30 second mark...
what the crap?
so finally i think to myself... what if i go lower... lower voltage...
vcore 1.45 + 2.1 vdimm (less than stock rated) and this is what i get at 9 x 405...
http://www.johnshim.com/forums/bleed...blend9x405.jpg
what the figgity? i need less than stock to get blend stable???
argh... has this been my problem all this time? too much vcore and vdimm???
Why are you running a 9 multi?Quote:
Originally Posted by shimmishim
just trying to push the fsb.Quote:
Originally Posted by fhpchris
the chip is capable of >3.7 ghz.
9 x 405 isn't going to be limiting the chip.
i could drop it to 8 and test.
Dropping multis gave me very bad results...Quote:
Originally Posted by shimmishim
9*435 primed
7*320 would not even boot
The lower I made my multi, the worse it got...
ah, interesting!Quote:
Originally Posted by fhpchris
right now i'm testing to see how long 9x405 will go on blend test.
i'm about to go to sleep so we'll see if it's running in the morning.
thanks for the suggestion.
I've known this since the 1.2 bios. The board doesn't handle the lowering of multis well. It ocs best at the stock multi except if you have a X6800.
This is one of the things I wish they would fix. Shimmishim, how do you know your cpu is good to 3.7ghz? Did you previously have it in another board?
i've tested it at 3.75 ghz or 375 x 10 orthos blend and small fft on this board.Quote:
Originally Posted by MrDeeds
also blend error'ed out last night at 1 hour 7 minutes :) but i'm happy since i've never been able to make it past 18 minutes...
i'm gonna keep tweaking away! but now i have a strange feeling that my ram is to blame for all my troubles...
I bet it is the board, I dont even run blend on this board for that reason...
so how do you test your ram?Quote:
Originally Posted by fhpchris
I like this board. For some reason it's getting me about 150Mhz more from my 6700 than it could do on my Badaxe1. Last night I gamed BF2142 for a little over 3 hours at 3800Mhz. 3650 was the last game stable spot on the badaxe1.
I'm glad I gave this board a chance, after the last 3 Abit boards I tried being disappointing it's nice to see Abit can still make a good board.
Use memtest. It will allow you to isolate the ram and see if its the issue. If you don't want to sit through it I think tests 5 and 8 are the best for isolating the issue. From my experience, test 5 errors can be solved with additional voltage.Quote:
Originally Posted by shimmishim
AB9 doesn't work with lower multis either with a e6600.:confused:Quote:
Originally Posted by MrDeeds
memtest is no good for detecting errors imo, i can clock much higher in memtest then i can in windows.Quote:
Originally Posted by MrDeeds
a pi 32m is usually stable enough for me, but sometimes your cpu can run at higher mhz in pi then in stress apps, so to me the best performance test is running 3dmark05 cpu benchmark in high prority and a 32m....
if niether of those fail your system is stable, IMO.
prime or dual 32m :)Quote:
Originally Posted by shimmishim
hi just got my abit aw9d-max board installed and ready, but i'm not familiar with the bios so i wonde if someone can tell me what to enable and disable in bios to get as hig clock as possible rest of my hw in sign, yhanskfull for answer...
This BIOS is layed out just about the same as any other board out there. Disable ur extra CPU options, drop to 1:1 divider, 1066 strap, pin vmch @ 2.0v, clock balls off of fsb (as far as possible with this board at least...).Quote:
Originally Posted by chefnr1
good luck with the FSB you'll barely get over 400 orthos stable...
max i've booted is 420 and clockgen'ed to 440 to do 1M run.
Frag - you the same Frag that got banned from the Abit forums? hehehhe.
ok thanks for the answer will test that :)
shimmishim this abit board will be fully v-modded so i can expect som better fsb...
Go at it chef! I wanna se 460+ 3d benchable by tomorrow! :)
how far you planning on pushin' the vmch?Quote:
Originally Posted by chefnr1
the only things that'll help you are the vmch and vfsb ... at least from what i understand
vmch and vfsb i have to test out how high i can go.
zeeke i will show you but not tomorrow must test it out first it's newly installed and i shall tweak down OS and optimize it first, and v-mod and do some cooling mod on gfx then i'm ready to go for it..
and offcourse a cooling sollution for nb too
Well Vfsb is a bit tricky on this one, I usually ran 1.42V (real) for high FSB but a couple of days ago I dropped down a bit and found the sweet spot on 1.35V So play around a bit with it.
The most importent one though is Vmch, think ive maxed 2.03V but it doesnt really do anything over 1.98 or so, I just cant hit a single mhz over FSB 473
I have something of note that may be helpful. A couple weeks ago I was able to do around 480x9 to 485x9 pretty much at will. This was with 2-512mb sticks or ddr1000. Then I switched to 2-1gb sticks of 667 to see about shaving some time from a 32m pi. Well it did not clock so well and after giving up I had forgot about changing rams. Since then I have been pulling my hair out to see why I could not get back to anything close to where it had been. Even changed cpus and went and bought a "B" today to try. After giving it one more try before changing processors again, I changed back to the higher rated rams. Yes, I know, pretty swift. What I learned though is that rams that run much higher at a 2:3 ratio, even with lower volts, will not do the same at 1:1 even with more volts. I could not get a 32m to go even past the first iteration at ~450 with alot of juice. But at 2:3 they would do 490 w/2.44v easily. For some reason 1:1 puts more demand on them, imo, even if they can run higher (and you know it). The same may be true for being able to run them higher on other boards as well. Screens of ram running at super high speeds with a 2:3 or other ratio may not bode so well on this board at 1:1. I think that may be the factor in some cases of not being able to obtain as high a clock as some would like. I will try to get some ram that is of 2-1gb sticks that will do 500 and test further. Below is a screen with 490 2-gb at 2:3. Fact is that these would never come close to that @ 1:1. Is this the norm and I was just not aware? Anyway I'm glad it still works good with the original rams and will try the "B" cpu tonight. EDIT** It could be a limitation of the board though with high fsb. But if that was the case, then I should not be able to run higher with different rams. (thinking to self)EDIT** Working back towards previous.
Yeah, abit forum fanboys are nubs. I couldn't find a lick of useful info there so I told em all to piss off :rolleyes:Quote:
Originally Posted by shimmishim
Quite the mature dude.:rolleyes:Quote:
Originally Posted by FragTek
New bios out 13.04B
http://forum.abit-usa.com/showpost.p...&postcount=198
dont forgot, use it at your own risk
BIOS works ok... Dropped my max stable clocks from 446x9 1:1 to 425x9 1:1 however :( Reverted back to earlier beta.Quote:
Originally Posted by Hilfiger
oh again a new bios.. im surprised :I
what date is it? and some changes?
Anybody try a Kentsfield with the new beta bios? I can't get anything beyond 295 Mhz fsb stable with any of the previous ones, and raising the multiplier only works on 2/4 cores.... Guess I'll have to try it myself... :p:
11/23, so it's younger than the B03 by just one day. Changelog is still unknown. Users on a german forum found that they changed the thermal monitor options once again, now it's just "enable" or "disable" instead of choosing TM1, 2 or disable.Quote:
Originally Posted by Virtual
Yup, that's the only change I can see as well.Quote:
Originally Posted by NicoOCZ
No difference with Kentsfield unfortunately. Not even 290 MHz fsb will run orthos blend (fails within seconds), and still only 2 of 4 cores react on increased multiplier :(
ty for the info.. so i dont think its worth to flash :(
IMHO, no it's not worth it by any means. Too bad though, I suspect I'd be able to make some pretty nice benches as soon as the multiplier issue is sorted...Quote:
Originally Posted by Virtual
I replaced the sp2004 screen with this one. Same voltages except for mch.: vcore:1.416 vdimm:2.112 mch:1.87 vfsb:1.465 [IMG]http://img207.imageshack.us/img207/5...1x9141vka9.jpg[/IMG]
Quote:
Originally Posted by coop
um... your screenshot is uhh... interesting?
in 5 hours, you should have done more than 3 iterations...
this is just a 20 minute run...
http://www.johnshim.com/forums/bleed...blend9x405.jpg
Ha, I was kind of wondering wtf..but not used to doing those so just was not sure. I will try again and replace previous screen later tonight. Ran a 1m before shutting off. 440x9 w/1.41v. This is the "B" cpu I got yesterday. 631B057. The vid is 1.25 compared to 1.2875 for my other one. Don't really know what that means but this one seems good too.
@coop - try orthos...
i've noticed that sp2004 has this problem where the second instance started will get stuck on the first iteration...
for whatever reason the 1st instance of sp2004 gets allocated all the memory and the second one doesn't...
one thing you can try if you don't want to use orthos is start the second instance of sp2004 when the 1st instance hits the 8K FFT length.
I see, I have tried assigning each to a cpu also and the same thing happens. I will try the 8 fft idea.
edit: frak, how long will that take?
about 15 minutes :)Quote:
Originally Posted by coop
Maybe a stupid question, but is that a hardware mod?Quote:
vfsb:1.465
RLM
^^ yes it is a readout that was provided from a post by crotale. http://www.xtremesystems.org/forums/...3&postcount=22
@shimmishim, I could not get sp2004 to go in any config. and didn't feel like running orthos for 4hr and 18min only to have it quit for whatever reason.(if it would have ran that long). Dual 32m works for me. I replaced the post above. Note< the temps in core temp are wrong. Sometimes it takes them a couple second or so to be correct, 7c to 15c in there somewhere.
Guys, I need help=(
I've read all the topic... but still haven't desided to buy the Abit AW9D-Max.
I have the money, but I'm still thinking of the alternatives.
I'm going to use it with Pentium 524 first, then in spring I'm gonna buy a E4300/E4400 or maybe the new Xeon 3050(smthn' else??). The Question is: is the MoBo good enogth or must I be looking for smthn' esle? Maybe on the P965 like the Asus P5B-Deluxe?
I need a stable MB for 24/7 with OC, good for C2D (Xeon), good for OCing em...
Gosh... I've read so many reviews: Abit gets really High makrs...
Can I order the MB and be calm? Can I rely on the new bioses that'll be made in future? Will I get 450+ fsb? (I know that the questions may sound dumb, but I need youre IMOs :)) Or must I look for smthn' else (then please name it)?
for any 975x board, most people would recommend the e6600 or higherQuote:
Originally Posted by iSlash
for any 965 board, most people would recommend the e6400 or lower
the reason why is because if you want to obtain a high overclock on the e6400, let's say 3.6 ghz, you're going to need a board that can run 8 x 450 which is possible on a board such as the gigabyte ds3 or asus p5b-deluxe.
shimmishim
10x :) Not the 1st day in OCing :) Thats why I've written about the E4300/E4400 - they'll be anounsed on the 21/01/07 aprox. and'll have 9x/10x so that'll be just fine :)
The question is about the MB - is it good enogth for it's price? Should I buy it and wait for new bioses? Or is there smthn' better???
+ people are doing 445+ with the bios new betas - that'll do for me even on 8x %)
great!Quote:
Originally Posted by iSlash
i think it's worth it.
i was going to get rid of this board but after testing out a asus p5b-deluxe, i just liked the feel of the abit board better and it's more o/c friendly IMHO.
if the 43/44 will have a 9x 10x mult then i say this board is a good fit then.
as for the 445+ with the new bios... not too sure about that...
the max i've been able to get out of my board is around 440 but that's only 1M stable...
i'm gonna be getting a 3060 here tomorrow so hopefully with the 9x default mult, i'll be able to push it.
shimmishim
So why di you think the Abit AW9D-Max is better than the Asus P5B-Deluxe?
I've been thinkin' of the 2nd too...
is not, simple as that, if you dont run sli go for asusQuote:
Originally Posted by iSlash
I like mine. Granted my board is not stock so my opinion is not so relevent. However the uguru feature is very nice compared to rebooting all the time. Plus having higher multi's available would make alot of difference on this board, see first post on this page for example.
10x guys.. I'll be thinkin'...
Actually its very simple, and if you've had a P5B Deluxe you might feel differently. I would recommend the P5B-D only if you need up to and over 500 fsb. Otherwise your going to need a soldering iron and a couple vrs to get it in the same realm with the aw9d-max. Dont get me wrong if you have the mod skills then i think the P5B-E or P5B-D will give you the best all around overclocking compatibility no matter which c2d you use but the aw9d-max is pretty good and hopefully will get better once we get some more bios updates.Quote:
Originally Posted by Hilfiger
Quote:
Originally Posted by MrDeeds
not everyone have phase, soldering and volts mod are not good for air
p5b does all that on air whithout modding
I'm basing my advice on air results and both boards were unmodded. The P5Bs vdroop is one of if not the worst i've personally experienced with a board. At 1.6v in the bios it droops to 1.56 in windows and that goes to like 1.53 under load.
It would be nice if everyone had a cpu that does well with low volts but i dont and that was one of my main issues with the board. Other than that you'll also have to understand i had it when it was cold-booting which they have now fixed. So at the time i didnt have such a great experience with the board and that was from my own experience not just from seeing how well others were doing with it. The board wouldnt boot higher than 387X9 for me and my highest boot on air with the AW9D-Max was 412X9, i wasnt stable at that fsb because of my cpus limitations but i guess i was sold at that point on the abit being the better board for 9x and up c2ds.
I dont see why your always putting down the board hilfiger, when i look at all the other 975X board we're all hovering around 450max fsb without mods and thats what the board gives. The dfi infinity board is the only one that seems to be pushing past the 450s consistently but look how long the boards been out. The only 680i boards better than 975X are the asus ones and they werent even available when the majority of us got this board. So i really dont know what your looking for here. fhpchris posts great air results with the board all the time and i dont believe his board is modded in any way other than a swiftech nb cooler. I dont have a problem with you liking the P5B-D over the AW9D-Max but what are you basing that stance on?
Well, found out that the AW9D isnt locking the PCIE bus, all my x1950s and my 8800gtx work fine!Quote:
Originally Posted by shimmishim
I plug in a x800 vanilla, and it cant even run stock speeds @ 434 fsb, yet runs fine @ 266*9.
I would think that anyone that is using SATA on this board(or a crappy ide hdd), or VGA that is very sensitive to PCIE clock speed would have problems!
I run raid on the Southbridge with 4 drives, so it seems that ~435 fsb is possible with that setup ;)
I bet everyone that is ~445+ is really running only 1m or using a HDD that works well @ high pcie speeds
I get the gist of that, but am a bit ignorant when it comes to PCIE, drives, raid etc. I just have a 250 gb drive connected into Sata 1. I forget the manufacturer.Quote:
Well, found out that the AW9D isnt locking the PCIE bus, all my x1950s and my 8800gtx work fine!
I plug in a x800 vanilla, and it cant even run stock speeds @ 434 fsb, yet runs fine @ 266*9.
I would think that anyone that is using SATA on this board(or a crappy ide hdd), or VGA that is very sensitive to PCIE clock speed would have problems!
I run raid on the Southbridge with 4 drives, so it seems that ~435 fsb is possible with that setup
I bet everyone that is ~445+ is really running only 1m or using a HDD that works well @ high pcie speeds
So what's the score then? Is there something to look out for when buying a drive? Does it affect other ide devices like dvd drives etc.
RLM
How are you guys determining the vfsb voltages you're running - readouts from the mods on the physical board itself? Thanks.Quote:
Originally Posted by zeeke
uguru reads the vfsb voltage. they call it cpu vtt 1.2V.Quote:
Originally Posted by Zucker2k
the difference between uguru and the actual measuring point is 0.02 volts.
uguru reads it 0.02 volts higher.
DMM- readout from physical board.Quote:
Originally Posted by Zucker2k
Soo, what's the catch ?Quote:
Originally Posted by fhpchris
How did you figure it out regarding the PCIe lock ? :confused:
You sure ?
In case it hasn't been posted yet, changelog of 13 B04:
Quote:
a. BIOS version 13.B04 has released for solving CMOS date lost all data back to "fail safe mode" issue.
b. Latest OC status: http://valid.x86-secret.com/show_oc?id=132570
because i gived away my pb5 for this :DQuote:
Originally Posted by MrDeeds
Good news.........A birdie told about 1333 strap......;) :D
:woot:Quote:
Originally Posted by hipro5
thanks hipro!
and i'm so glad i didn't get rid of this board (assuming this is true).
mh.. 1333 strap.. mhh :rocker: :D
they sayed this long time ago, just on release, any idea when bios for it?Quote:
Originally Posted by hipro5
That birdie deserves a reward ! :banana: :banana:
:clap:
WoW! If we get 1333 strap u can bet ur arse i'm gonna keep this board... Did the birdie happen to mention anything about a possible release date? ;)
if no bios gets released within this week, we get to bbq the birdie and then feed it to the worms
got some new memory and a new chip...
figured i'd try some new hardware out to make sure it was the board causing the board to not be able to blend @ 400+
so i tried the mem and the cpu and still can't do blend.
so i've given up on it... for now.
it's still a fast board and i can now easily boot 420 with no tweaks where as before with my 6700, it'd take a while to get 420 to boot.
That's great news. Any news about whether it will have "proper" Kentsfield support?Quote:
Originally Posted by hipro5
You didn't get it!......That birdie not ONLY told me that newer bios with 1333 strap is going to be released!......He has already send me the bios too.....:D
Though it's a preleleased ALFA version for the time been.....
It HAS the 1333 strap but unfortunately not above 5:4 rams.....I mean by that, that the rams - for the time been - are underclocked so as firstly to test the 1333 strap and then see what will happen.......I'll test the bios today and I'll report back... :)
:DQuote:
Originally Posted by situman
Can you also test to see if the pcie bus is locking correctly?Quote:
Originally Posted by hipro5
i´m sure he will ;)
I'm pretty sure the birdie just forgot to tell you, but I'm on that list to get that BIOS, mind forwarding it to me? :pQuote:
Originally Posted by hipro5
Sorry, I just can't for the time been.......:(Quote:
Originally Posted by FragTek
EDIT:
http://www.thelab.gr/images/Hipro5/T...1333_strap.jpg
:woot: :woot: :woot: :woot: :woot: :woot: :woot:Quote:
Originally Posted by hipro5
Please request to add support for FSB vTT adjustment up to 1.8v (currently 1.2v only). I'm too lazy to vMod a board nowadays, so I hope they will add that option to help Kentsfield with better FSB.
only Alpa?? So that means it will be AFTER the next bios release? We've had 3 beta bios already. Now we are back to Alpa, that's nice.
lol. for real.Quote:
Originally Posted by situman
just release official 13!!! :stick:
so we can say there is now an alpha 14 available :)
...so whats up Hipro? May we test it too pleeeeaaaase ? :stick:Quote:
Originally Posted by hipro5
I´ve allready done your vfsb- and your mch-booster-mod but I´m still maxing out at 386FSB. I´ve tested lots of HQ-RAMs such as Cellshock, G.SKill, Geil, Team so I reeeeaaaaalllllyyyyy neeeeeeeeed that 1333 Strap. :banana: :banana:
Greetz!
To the powers that be, we've suffered a long time with this board. Help us get rid of some of the headache. And by this I don't only mean Hipro5 - we don't want Hipro5 taking the heat for any leaks. However, if there are any XS members who have access to the new 333 strap bios please do the community a favor and give us a hint on how to get it. I'd recommend that we all go hush hush on this one. We've spent too many hours on this one, trying to make it stable, give us the opportunity to try something else. In fact, I wouldn't mind at all if someone were to surprise me with a PM :D. Go team XS!Quote:
Originally Posted by Zucker2k
I could care less about getting it right now, i just want to know if its worth waiting on. You guys shouldnt try to pressure hipro or anybody else to make it available because once he does that the first thing some idiot is going to do is post it on rapidshare for the world.
I just want results. Please tell me the board is doing over 500fsb.