new one :peace:
i think i'm quite happy with this:
http://www.abload.de/thumb/818ghzcnqp4y.jpg
nb vid 1,325v :up:
new one :peace:
i think i'm quite happy with this:
http://www.abload.de/thumb/818ghzcnqp4y.jpg
nb vid 1,325v :up:
HT ref =266 it's a nice number;) :up:Quote:
Originally Posted by Oese
mmh unfortunately 266mhz is not stable with cnq (and thus with stock nb multi), but 265 is :D not bad as well, especially the ht clock is extreme i think. 2380mhz ht-link :shocked: but it is also bench and gaming stable, when the gpu traffic kicks in.. knock on wood.. ;)
guys pls help!
since today, both my lan-ports on the m2r do not work anymore. i cleared cmos, flashed bios 3 times, went on optimized defaults none work.
they do not show up in the hardware manager and the lights are not flashing neither on the router nor on the board side.
i now use a pci lan card which works, so, its not router, cable or anything.. can you imagine a possible cause? i know if too many pcie-lanes are occupied this can happen, but thats not the case. all settings regarding this were left untouched.
it happened after trying to x-fire a 3870 with my new 4850 (which does not work although some people reported about this possibility on the net) and doing some driver deinstall/reinstall. Then i realized theres no net and since then i tried to get it back to life again.
you can think of some things left to try? it is definitely no windows/vista thing because the lights on the nic's arent flashing during boot up like they should and the nics dont show up in the post-irq-table as well (i dont know if they were included in that table before.. could anybody perhaps check this
out?)
that was what i needed for a completely ruined sunday afternoon :shakes:
Not any help to you but one of mine died when the board was only 3 weeks old, I'm holding out to buy a new board with SB750 before I RMA this one, there is no rhyme or reason why it failed on my board it just vanished from the system totally, posibly some inherent weakness in the board fabrication.
Olli ....whats the common thread with problems associated with Phenom OC that everyone seems to run into? Phenom High HT OC is the root of all of the problems that I have seen to date and no one will listen when I say that Pushing high HT OC with Phenom on these boards is a big NO.. NO:( but..... No One listens:rolleyes:. Sorry brother I think you boinked your board from pushing to hard:(Quote:
Originally Posted by Oese
If you are Lucky........ Clear CMOS with the battery out for about a Hour then with the cmos battery in the Upside down position in the battery socket leave it there for about 10 minutes then put your jumper and battery in normal position with only a PS2 Keyboard plugged into the back Boot up Press F1 continue to desktop (unless you need to set the Date because of unactivated windows copy) procede to desktop. See if that works if No Go Reinstall WINBLOWS VISTA !
I would suggest that you heed my warning about high HT Associated Overclock with the First Phenoms and Phenom Capable Motherboards.........Remember ... Growing Pains its a NEW PLATFORM!!!!!!
mmh but why the lanports?
and really, if i cannot oc this thing via htt i cannot oc it at all.. i think its not the right way to limit oc to the black edition processors.
and what else.. board is running at stock more or less. if the phenom broke, id agree, but did i uppen one voltage on the board or somewhat.. no, really... and the phenom does not run at high voltage as well.. 1,38v&1,325v cannot be too much.. then i will have to change to the newer sb750 boards.. this phenom ocs too good to give up on this...
what could be the cause was trying a pcie clock of 120mhz.. i ever feared uppen the pcie but since i read that much about it, i tried it out yesterday since today.. now thats stock again. maybe that broke the lan-controllers as thats directly connected to them.. thats doing me a harm :eh:
really campbell, i think the board is pretty much powerful enough to power a phenom at any given speeds... sure some phenoms were not able to and especially high voltage can pose a thread..
but i dont know the lan ports.. maybe my feeling was right better not to uppen the pcie.. well now i did and i paid for..
i will try one hour of cmos clear whatever and then cmos battery upside down..
Sorry to ear that.Quote:
Originally Posted by Oese
Can't occur me anything else, have you try to enable both networks ROM in bios? Just to check if they boot.
Now, you got a good excuse to by a new SB750 board.:p:
network ROM? dont know will have to look after it.. did not find such option in my run to get rid of the problem..
sb750 would be ok if there were any boards available, particularly by dfi :eh:
I had 124 in PCIe w/ 3870 CF, didn't had any problems related.Quote:
Originally Posted by Oese
I thinking in buying an sb750 board, but this time I will avoid DFI, unless if it have very good feedback.
Quote:
Originally Posted by Oese
Sorry to mess, I checked DFI manual and it didn't have that option to enable Boot ROM from the LAN controlers.
Like I said Olli AMD is going through growing pains with their new Platform and like anything else it takes time to sort things out and get them to where you want them:yepp: As far as the D.F.I Motherboard is concerned just like any board maker they can only build to spec set fourth from AMD so its not a D.F.I thing as some would suggest;) I have already stated this that you need to clock CPU , NB , HT, Ram independently then use HT to fine tune the last of you're Base overclock:yepp:Quote:
Originally Posted by Oese
yes i know i listened to your arguments but i have a 9750.. and that way i cannot overclock the cpu without ht ref clock.
perhaps i should have bought an 9850.. but.. that f**in 9750 just clocks too good!!
and i'm going through growing pains as well.. campbell, what should i say.. i believe in amd and thats not their AND not dfi's fault. the pcie clock i should not have upped it was i believe.. that phenom i have is one of the newer ones that can stand high ht i'm convinced.. thats just normal progression of the process..
Hmm I set pcie to 120MHz back in the past and ran it for a week or so without problems. :shrug:Quote:
Originally Posted by Oese
I used high ref HT's since day one with a 9500 whom died on day one in the dfi board due to dimm overvolting by accident (no access to sources here so i can not proofe). I also used high ref HT'son the 9600BE whom still works. The 9850BE also worked fine with high ref HT's but died from dimm overvolting (this time proovable by source code ,cuz it happend under linux).
With the 9950BE i only tested higher ref HT's for a day till now.
BTW: I did alot of thermal testings with the 9950BE. I wrote a script which modifies the cpu multis and voltages via msr registers under linux.
It then waits ten minutes and then runs stresscpu2 for 20mins. I used CnQ with 750MHz/0.9V for p-state-1 to get better cooling in idle.
I ran
CPU-VID / Frequency
1.475V / 3.1-1.6 GHz
1.450V / 3.1-1.6 GHz
1.425V / 3.0-1.6 GHz
1.400V / 3.0-1.6 GHz
1.375V / 3.0-1.6 GHz
1.350V / 2.9-1.6 GHz
1.325V / 2.9-1.6 GHz
1.300V / 2.8-1.6 GHz
Temperature diagramms look like this:
http://www.abload.de/thumb/t-1425rml.png
I used the difference between load and idle temps for y because the idle temp always was ~3°C below my ambients and those changed over the four days the test ran. x is the cou-vid used.
http://www.abload.de/thumb/temperaturesytt.jpg
This diagramm cleary shows the odd thermal behaviour this chip gets above ~ 1.375V with higher frequencies. The temps jump about 5°C and then they continue to increase with 1-2°C per 100MHz voltage increase or per 0.025V increase.
To make sure such a jump does not occure during uptime (which would result in a freeze) I need above 1.475V cpu vid. But with those voltages the system tends to power off after half an hour or so running the universe suite under linux 64bit.
For the record i ran the test also with air cooling:
1.3V with H2O: 2.8GHz 50°C 1.6GHz 40°C
http://www.abload.de/thumb/t-130077h.png
1.3V with stock cooler: 2.7 GHz 74°C 1.6GHz 58°C
http://www.abload.de/thumb/t-1300kla.png
1.3V with BigTyphoon: 2.8GHz 62°C 1.6GHz 47°C
http://www.abload.de/thumb/t-1300oci.png
1.3V with XIG: 2.8GHz 74°C 1.6GHz 57°C
http://www.abload.de/thumb/t-13000i7.png
My XIG must have some sort of damage because it was ~5°C better than the BT when i received that cooler and now it is on the same level with the stock cooler.
I tried to find a setting where 18A for the two 12V rails would be reached because that whould be the limit for a one 12V rail mobo.
1.5V and 3.2GHz
http://www.abload.de/thumb/3.2-1.59q8.jpg
interesting!
this p-tuner thing works with every psu? or is it mobo-related? would be nice to measure here also :D
very interesting and detailed testing! but no clue about that temperature increase at > 2,7ghz and ~1,3v, right? really weird..
dfi europe wrote back to me i should try a win xp32 install to see how it works out that way.. i'll do now :D
ok no luck and already prepared for rma. have the dfi cfx in now again.. superb feeling :D paired a opteron 170 @ 3ghz with a 4850 :D if that has been possible last year ^^ :D
Olli, was just popping in to see if you got the problem sorted out. Guess not so I see:(
no. i tried today after the board was out of the case and unplugged from everything and cmos cleared for 24h.
no luck.
but the dfi europe guy already has my rma request i'm just waiting for the number.. and perhaps they get my broken nb fan sense and control back to work as well.. that thing didnt work from the first day i got the board.. well second hand.. but i did not want to do a rma for that although working but not controllable fan header.
whatever campbell did you ever have a 3200cfx-drg? best s939 board out there except some bugs.. but lucky me i never had any problem with that feared freeze bug. well, as the board got old now the second pcie-slot is not working and out of warranty, but with only one card really works supreme still and my opti now only needs 1,456v for the 3ghz instead of 1,48v before, obviously thanks to my now polished cooler ;)
so i have fun whilst waiting for rma, and the hope of getting a completely working board back makes me happy in advance..
Any possibility these boards will support UB ECC?
With a Phenom?
With a 13xx Opti?
"With green neweggs and spam".
I ordered the ECC for my 4S quad before I realized they where unbuffered. DDR2-800 was too good to be true.
Haven't been up for quite a while
My board seems having trouble booting up to Windows XP
I tried Clean install but it still does the same, I even tried different hard drive and SATA cable
and having the same problem.
The second thing is the "SmartGuardian" no longer work , no matter how many times I reinstall / clean install the SmartGuardian comes with the Driver CD
It only shows "0V / 0 rpm" on every items
Finally the onboard Si3132 RAID / SATA Link "disappear" from POST from time to time
I guess my board is somehow broke
It still managed to boot up normally after a while.
Too bad the DFI reseller in Hong Kong isn't good and RMA would be No-No.
I was planning to get my hands on a HD4850 but now I have to start look for a new board for my rig
Good thing is , 790GX board is available in Hong Kong now , for now only Biostar and Jetway available
(Most of the dealer focus on selling Intel platform board more than AMD )
Just hope the board can hold on til more 790FX + SB750 / 790GX + SB750 release
DFI Europe told me recently that DFI sb750 boards will appear at the end of august..
Hey Oese don't know if ur LAN works
but the PCIe (3?) should be set to 2 and not 4 in BIOS otherwise
there isn't enough lanes.got a very limited Inet access atm:(
jep i know that pcie thing. checked this .. no luck unfortunately, was at pcie x2 and after cmos clear it should be set back to pcie x2 whatever..
but thx ;)
board is already send in for a rma. i hope they'll process it fast, but i think so. usually, a week or one and a half..
sometimes the only way to get the lans back is with an OS reinstall (this is assuming the circuit is not fried), however, it more than likely will only be a temporary solution as whatever caused the intitial failure is a board function and not an OS function.....
you would think that after several years of this problem cropping up here and there (actually, a lot) that someone would fix it....guess getting a $10 PCI NIC is the best you can do for now...
laterzzzz...................
jepp i tried reinstall and xp32bit install and its broken obviously.
whatever i have a pci nic, but i wanted to get all those board functions back to life.. so now its on his way to dfi nederlands..
P-Tuner only works with Gigabyte Odin PSU's whom have an USB connector for sensor output. The PSU is not the best efficiency wise but very cool for measurements.Quote:
Originally Posted by Oese
I have no clue if it's a cpu/pwm or water cooling setup thing. So I'll wait for a sb750 board first. At least I have a bunch of data to compare if the cpu should degrade in the future. :rolleyes:
Have you tried to reinstall the SB600 drivers?Quote:
Originally Posted by imamage
In Hong Kong i'd buy the Onda board.Quote:
Originally Posted by imamage
http://product.pcpop.com/000115922/Picture.html
Cool thanks for the info.Quote:
Originally Posted by Oese