I'd say that's a little high for water! What block and pump are you using ...?
Printable View
I'd say that's a little high for water! What block and pump are you using ...?
i am using everything that come with the thermaltake bigwater 735. a thermaltake aquabay M2 and a Swiftech MCR220-QP Radiator as my 2 other radiators in a lian li pc a71
there a couple air coolers that do better the the Big Water setup...I saw a review somewhere comparing them.. they had temp results etc.. I'll see if I can find the link I looked at..
Did you guys see there is going to be one more revision to the 65nm Phenom? Part of AMD's Continuous Transistor Improvement, or CTI. Probably Fall time..
That is interesting, wonder what will change?Quote:
Originally Posted by charged3800z24
On another note....
I just got a new case and in the old case I was running 2900MHz with default VCore (1.296 I think) and it ran great, no problems. Once I put everything in my new case everything ran cooler and I was happy......then BAM! BSOD. It did that three times until I upped my VCore to 1.376. Now it is stable. I cannot for the life of me figure out why it was necessary to up the voltage unless the chip has a sweet spot as far as temps are concerned. Anyone else have any ideas.
Hi
Ben Brown with the new case you are running dose it have more caes fan than before ? If yes maybe the fans are taking volts way from the cpu some how and Putting a biger load on the PS .
Hi
Well that the only thing i could think of that you change for one case to the other.The 9850be suck down juice like a pig at stock
clocks let alone when you over clock it. You take that and the 3 video cards that you are running that thing under load is going to be a power hog. Maybe adding the fans and if you got them plug into the motherboard was was all it took to change the power blance on
the board . Grats on the 3dmark-06 score that a great score for a all AMD rig.
Well you may just be on to something.Quote:
Originally Posted by merkk1
I tried to do some more 3DMark06 benching last night at the same clocks as my previous high score but it locked up pretty quick. I lowered the clock on Core 0 100MHz and it still locked up just before it started the last SM 3.0 test. I then proceeded to add an additional fan and it rebooted as soon as I plugged it in. So I then unhooked three fans and left the case open with a fan blowing on it to keep it cool and it completed a bench (another 17k+ run :) ).
So you must be correct about the power supply not having enough juice to run everything. Man I thought when building this system that 750W would be enough. Anyway I took my Thermaltake 450W out of my soon to be P4 folding rig and hooked it up to 1 of the video cards and two case fans. I will try and run 3DMark06 tonight and see what happens.
I need to see if I can find the maximum wattage for those cards to see if I can use the 450W PS to run all three, probably not though.
Thanks for the idea!
Hi
Ben Brown well dont feel bad dude i got a new power supply in my rig to and it a 750w ps also. In fact it has the same stats as yours
has and i got 3 case fan and cpu fan running off the mother board . It to is having trouble staying stable at 2.81 ghz . I dont
think it the 12v rails that are the trouble but the 5v rail that making the trouble . So to test what i think is wrong i try some things to night . I mover around the PCIe line on the video cards to see if useing the other one change how the stablely in the board was(i have 4 line to pick from ) i change over to the 2 6+2 PCIe line to run the 2 video cards and the hole rig way more unstable was crashing at
2.81ghz at stock v , 1.31, 1.32v before it would run stable for a long time at thoughts volts setting before . So i said screw this pull the 750w ps out ,now it only has a 28A 5 volt rail with a told of 180w for the 3.3 + 5v combine . I put my 600w ps back in it has a
55A 5 volt rail and the tag on the ps says 3.3v 30A + 5v 55A combine 320w .I power the rig back up at 2.81 ghz running stock volts
no crashing now for a couple hours going to take it to work tomarow run strest test and some other stuff for 8 hours at work see what happens.I thinking the 5v rail in the 750 ps can not handle the 9850be cpu over clock. I was talking with my boss down at work and
he think the same thing so we see what happens.
Well I hooked up another power supply to the pc last night and used it to run one of the three video cards and two case fans. I was able to push the phenom farther (3.3, 3.1, 2.9, 3.2 and 1.504 vcore) than I have before and it ran three 3DMark06 runs in a row. It did crash on the fourth test when I tried to up my video card clocks higher so I may try to hook up another video card to the extra PSU to see if there is still too much of a power draw or not.
Thanks for the help merkk1. Now I just need to find a god load balance for the two power supplies and pencil mod my video cards for more juice and break 18k. WOOHOO!!
Hi
Glad i was able to help and in helping you it made me rethink what i was going on on my pc and i was able to help my self so it was all good all the way around . The 600w PS is running everythink like a top the phenom clock at 2.81 stock volts ,2 HIS 3850 in crossfire,2 WD hard drives in raid-0, X-FI sound card ,4 gigs of DDR2-800 ,3 case fans one cpu fan,floppy and the LG dvd burner.Now what dose
make me mad is that new 750w ps i pay for can not do the same thing . I need to realy think over the replacement PW this time around . What every one i pick going to make sure it has at lease 250w on the combine 3.3v + 5v line because that ready what
is diff. form 750w ps . (600w got 30A 3.3v + 55A 5v for combine 320w , 750w 30A3.3v + 28A 5v for a combine 180w) .
I just got thinking today about how the nb has been clocking for most people. It seems to clock like the good old clawhammers and newcastles. My newcastle 3200+ only clocked from 2.2 up to 2.5 to benchmark, 2.4 stable. My NB clocks from 2.0 to 2.5 benchable, 2.42 stable. Not sure it really means anything, just my random thought of the day.
And for stable I'm talking load stable with rosetta/WCG running for 8 hours.
I noticed that too. Thanks for reminding me there was something else I'm remebering that may be important.Quote:
Originally Posted by Oldguy932
I have had voltge sweet spot.. but never paid attention to temps..
Any who.. Anyone notice the Phenom 9950 is coming soon.. after the launch of HD 4870. The Phenom 9950 is suppose to have a stock clock of 2.66ghz.. Strange. I wonder how its clocked? I wonder if the 9950 will have the new revision AMD was planning for the Agena core. I am eager to see.
it should be interesting what the 9950 can do but then if it does come out with a 10x266 or something along those lines that would give such a wierd HT link speed. The best i can get my 9850 to is 13.5, 13, 13, 12 @ 223HT. volage is an issue but anything after 1.387Vcore does nothing anything over 1.5vcore the system wont boot. anything over 2300MHZ on the northbridge the system crashes after loading vista. closest have have gotten to stable NB is 2200. he is a pic of how my system crashes everytime
http://img527.imageshack.us/img527/6...e017qo9.th.jpg
Calcifer, your NB may be a little too high...
But everytime my machine crashes like that it's cause my HT Link is too high.. :D
Lower the HT Multi by 1 and see if that fixes your problem, I get the feeling it might... ;)
Here are my settings Dave. HT clock 233, CPU multis 13.5, 13, 13, 12. CPU Vcore 1.387 (1.4 bios) NB multi 10, NB voltage 1.3V, HT multi 8, HT voltage 1.4V, Ram 932 5-5-5-18 2.0V GANGED (Crashes over 1000mhz). i lowered the ht multi cause i think tony got better overclocking with it at a lower speed. have tried down to a multi of 5 and doesn't make a diffence for me. the wierd thing is is that some time i can boot up from bios with all cores at 2.8Ghz be fine reboot to try some more and it wont make it to vista unles i take it back to defualt. i have pulled all my ram out and tried it with all the sticks 1 at a time and that doesn't change a thing... Overclocking phenom kills my inner child
If your vDIMM is 2.0v or higher you might want to nudge the Mem vDDQ a little. My RAM runs 2.1v (stock) and I couldn't run 200x15 until I upped the Mem vDDQ to 1.9v+ ...
how do i do that? there isn't an option in the bios to up it and if i change any voltage in AOD my comp freezes up. would it be called something special in the bios?
It's not in my BIOS either - AFAIK it's only accessible through AOD ...Quote:
Originally Posted by Calcifer
Some one asked what i thought of the may 5th beta on the DFI...check the pic for what I think ;)
8GB stable at 1066 3GHZ cpu...think im first to do this, heat and voltage play a major role here, memory timings by Auto except I forced 195ns tRFC, vdimm 2.0v and CPU/NB running 1.4V.
http://www.ocztechnologyforum.com/st...Z_8GB_1066.png
CoreDamage catching the odd spare cycle it could here, 4 instances of HCI memtest running each using 1700MB of ram, ( raja and Praz use this as its so brutal), full 100% coverage then I got bored ;)
Seems my 1066 spd is working well :)
its actually hidden in the DFI bios along with VTT, I may ask they unhide it if it would add stabilityQuote:
Originally Posted by QuietIce
Quote:
Originally Posted by Tony
Thanks Tony I appreciate it;) Have you tested DMM CPU/NB to see ow the far off the varibles are off on Bios read out?
It seemed to resolve the stability problems I was having. Before raising the Mem vDDQ I could get 3.0 GHz but only by running 188x16. Now 200x15 has been crunching 24/7 for over a week. :)Quote:
Originally Posted by Tony
I did lower the HT Link speed as well but I had tried that before and still couldn't get 200x15 to run stable until I changed the vDDQ ...