Printable View
i have warned everyone about how incredibly hot this NB gets.Quote:
Originally Posted by dissident
the one and only thing you need to do is buy a 40mm fan and hotglue it to the top of the heatsink.
i did that and my chipset never gets above 40c.
(i wonder when people will start taking me seriously about crap like this)
:YIPPIE:Quote:
Originally Posted by aldamon
yeah it was REALLY early in the morningQuote:
Originally Posted by Lestat
anyways
F3 bios on website...final one works best for me
Highest stable overclocks (ram related issues with F4C)
Better RAM overclocks
example 460*7 which is gizzillion hours prime/SPI stable on F3 will clonk out in F4C no matter what >>> reason is memory just craps out...........EVEN at 460MHz whereas this mother can do 500+ 4-4-4-12 2.2v 520+ 4-4-4-12 2.3v in F3
i was testing the Normal, Fast and Turbo settings in F4C only to have Normal and Turbo lock up half way and Fast setting on the second run locked up as well :nono:
after the resetting and all the usual caper including ONE bios change at a time >>> no go
back to F3
everything works...........RAM does 503Mhz 2.2v 4-4-4-12............I am just completing my testing for Normal, Fast and Turbo settings on F3........all 100% stable
dude my method worksQuote:
Originally Posted by Lestat
i even tested the CPU at 1.5v 460*7=3220MHz; MCH/FSB max volts and the temp didn't go beyond 38C reported and luke warm to touch.........i've described my method of cooling in another thread
also i've had two DS3 motherboards so far and same results........one other thing that kiwi mentioned regarding mosfets and other things around the CPU getting hot......once again.....luke warm here
method 80mm fan mounted above left side of the RAM sticks directly blowing at MCH area..............TTBT cools the rest...............no dramas
yeah i can agree it really doesnt take much to cool the mosfets.
but when there is no airflow.. .yeah they get pretty warm real fast.
520 thats a record man. no matter what multi thats a record.
what ram was was it you were using again ?
520 was RAM OC not FSB............Quote:
Originally Posted by Lestat
this board can do 520 though but i'm just not having luck with my mix of hardware
my RAM PC6400 G.skill HZ :)
here's a couple of screens from previous owner......he got up to 520 1M SPI....525 froze........he used some cheap Corsair value ram with 2.1v and Auto timings lol 1:1
http://img395.imageshack.us/img395/6...0515fsbeh1.jpg
http://www.lie.id.au/pics/500fsb_3dMark.jpg
ok guys i need you fellas to try something
do whatever you gotta do to boot at your MAX FSB (this is aimed at those of us who are stuck at 400 or 420mhz FSB)
now go into the bios and change it instantly to 435 FSB.
pci-e 100
voltages. set your ram to whatever will work for you ram
fsb mch and pci-e to max
during my testing im finding that there is sort of a big gap int he cpu vcore.
i can do 3.6ghz at 1.3375v BUT in order to go anyhigher stable i have to go right up to 1.4v PLUS this also allows the FSB to go higher.
now. im at 435mhz bootable from the bios and 1.43v vcore.
i am running memtest cuz i am using my crappy samsung but if that looks of i will boot into windows and try it out.
then i will switch to my gskill crap and see how well that handles it.
i am having a hard time doing much with the gskill as its a really wierd set of ram.
high volts low volts it always acts differently. i can never tell what to use. there isnt a set voltage or CL or SPD that is working.
try it guys see what happens.
once you hit 400, or 410 or 420 whatever. make the leap right up to 435. see what happens. and dont forget to increase your vcore accordingly
it didn't help meQuote:
Originally Posted by Lestat
i think i couple of important things in overclocking this board are
1) motherboard "bend"
2) RAM
3) vdimm
4) other voltage
it seems as though the signal traces on motherboard are overlapping when you add too much voltage along with high OCs
I really think that once there is a sub 1:1 divider we will see some monster FSB OCs
Where should i put aht 80mm fan? I can't envision it. Ironcially i was just out in the garage soldering a 3 pin header on a fan and was thinking of using it somewhere.
Here's a pic of my case. Sorry it's blurry, but can you help me place it?
http://www.xtremesystems.org/forums/...1&d=1154745379
lestat didn't think much of this but it WORKs and the fan is quiet at the same time
fan is closer to RAM/MCH now
http://img140.imageshack.us/img140/3743/dsc03086ov4.jpg
sweet thanks. What do you mean 'mobo bend'?
Lestat, when you get a chance, could you answer my question?Quote:
Originally Posted by aldamon
dinos
i think witht he 7x multi you are just at the max of the baord. period.
thats the highest i have seen anyone do yet with a 7x i believe.
do you have a 6600 or something ? id be VERY curious what you can do with a 9x multi or higher on that board.
your also at a 2000mhz bus
thats just gotta be making that chipset just bleeeeeed!!! know what i mean.
Quote:
Originally Posted by aldamon
thats a big yes my friend. no more PIO mode
but be sure to uninstall the gigabyte sata drivers.
if you have raid then dont of course lol but it should right away change over to DMA
if not just right click the 3rd primary ide controller. the one your cdroms are on and chose to uninstall. then reboot and come back in and see what happens.
i know :( someone unlock this damn CPU i want to hit 5GHz with it :DQuote:
Originally Posted by Lestat
single stage rotary in basement is getting bored heheh
Bios Bootable.
http://valid.x86-secret.com/show_oc?id=109867
i had to take a HUGE leap in vcore to reach this.
so... now this is just a preliminary thought as i need to do alot more testing.
vcore and fsb are sharing a direct relationship.
again thats a far fetched guess but i find it odd that 1.33v is 3.6ghz stable but i have to run 1.45X in order to 3.9ghz.
its only a 35mhz increase in FSB...
wierd but possible.
again much more testing needed so i'll do my best to bring more accurate results.
32m Pi coming next.
you know it was the same with AMDQuote:
Originally Posted by Lestat
when you start reaching the CPU silicone limit it need A LOT of vcore
nope 32m not stable at 1.45X i am at 1.49x and 32m is 1/2 done.
hhmm i dont like having to run 1.5v and not even hitting 4ghz yet.
i cant recall what the other guys where getting on their ES chips with something like the bad axe.
does 1.5v sound right for 3.9ghz ? i cant recall.
Quote:
Originally Posted by dinos22
yeah i know but whats wierd is the 400mhz@ auto mch fsb and pci-e @ 1.33v and then having to leap allll the way to 435 for it to be stable and then the near 0.20v increase to do that.
just seems a bit off. i cant put my finger on it other than a simple wall in the cpu.
i suggest you go in smaller increments from 410 upwards and see how the voltage scalesQuote:
Originally Posted by Lestat
ohhhh believe my why else would i suggest others use small jumps.
see this is where i say things like "i will do extensive testing" and " i want to ensure i bring you guys the most accurate info"
im not just sticking my d|c in the first hole with hair around it.
so here it is 435mhz fSB bootable and 32m Stable (now its time to see if its 3d stable)
32 stable for one instance :p:
nice going there man
which ram have you got in there now
this is the mediocre gskill 6400 ZX
its really not that great of ram for 6400.
i have no clue at all what chips it uses, and never even seen anyone else use it.
she wont even do 1000 5-6-6-x w/ 2.4v
Thank you! I uninstalled the Gigabyte drivers and all is well. One of my Primary IDE Channels was still showing PIO mode, but I changed it to DMA and rebooted. Now eveything is DMA again.Quote:
Originally Posted by Lestat
Stupid Gigabyte drivers indeed!
Are there any new ratios?