I dont know how hot it gets.
Printable View
I dont know how hot it gets.
Did some testing about the 200 FSB boot. It has been said that booting at 200FSB gives performance boost because you get full pat and so. I did 10 superpi runs with my system everything same between each series of 10 runs but one was booted at 217FBS and the second was booted at 200FSB and then increased to 217FSB with clockgen. Results were as follows:
FSB200boot average time: 28.713
FSB217boot average time: 28.714
It seems there isn't any performance increase there...waiting for others to test.
PD:YEs hyperpath 2 is enabled in bios and spread spectrum is disabled.
hey its 1/100th second faster! :D
so this is just a myth? hmm
I "just" killed 2 sticks and not 3.
The two stick burned in a similar situation... after a crash because trying to overclock too much for them.
I allways have cpu-z and clockgen together open and sometimes one or another doesn't report the exact settings for that moment. Sometimes have to restart cpu-z and/or click on clockgen in the get setup (or something like that) button to refresh info.
In the two situations I raised the voltage to see if perhaps the crash gones (while running).
What I mean with 915 tweaker is that if it's possible that it uses some instruction that this board doesn't like or can't handle somehow...
Yes, it seems it is. Have to try if it gives better overclocks or better stability when overclocking.Quote:
Originally Posted by saaya
ratax, i dont think 915 tweaker is the cause of this, you were at high memory speeds and then tried to set timings to -2-2 right?
my guess is the memory didnt run -2-2 at those speeds and thats why the system crashed.
why the memory died is a mystery though...
Ok, yesterday me, ojdr and RickY puted the Dothan 780ES+P4GD1 under cascade but we got some weird problems.
At multi 17 we could bench SuperPi @ 3700Mhzs (cascade was only doing -75, normal is -100), so far so good.
Then we dropped multi to 12 to check max memory frequency and we got the same that we did on air, 275Mhzs 2-2-2-5 out of BH5.
Then we tried with multi's 13 to 16 and no go at all. Multi 13 we could only get 13*260, multi 14 we could only get 14*245, multi 15 we could only get 15*230 and multi 16 not even 220 FSB. So the higher the multi less FSB we could get without getting to the max the CPU can do.
Anyone with the same problem? Or any ideia on how to solve it?
Have you made any chipset mod on it? Which PCI-E Freq are you running at?
Chipset mod is done, we tried from 1.71v to 1.85v, no change.Quote:
Originally Posted by NiCKE^
PCI-Ex is set between 125 and 135, no go at any value. Also booting at 170+ FSB to have locks working.
Btw, dont know if anyone reported this before, temperature reading is bugged bellow -22ºC, it always reads -122ºC with less than -22ºC real.
thats a weird problem... sounds like a cold bug on dothan LOL
could it be that the head isnt making good contact with the cpu?
The contact was exelent, but even if it was bad, with multi 17 the CPU benched @ 3700Mhzs, should do the same with lower multis as long as the board does the FSB needed (and it does, at least with multi 12).Quote:
Originally Posted by saaya
hmmmmmm maybe the multis didnt work correctly?
maybe they didnt apply?
i heard that on the aopen 479 915p board the multis look like they get aplied but they arent.
12x275 3300
13x260 3380
14x245 3430
15x230 3450
16x215 3450
hmmmm coincidence?
the cpu speed is always around the same...
i benched on my rig with banias and dothan and the multis applied correctly though... lower multi higher fsb same cpu speed was a higher score... weird...
Multis were applied for sure.Quote:
Originally Posted by saaya
I've also noticed that coincidence ins CPU Mhzs, but I have no ideia on why that happens.
I'll try some dry ice soon to check if the problem persists, if it persists I'll try a P4C800 if it continues I have no ideia on what to do :S
Anybody have an easier Vcore mod for the P4GD1 than Kingpin's as show here:
http://pic8.picturetrail.com/VOL256/.../120540761.jpg
It's pretty tight in there. I rather solder to a cap than try to hookup to that pin.
you just have to trace the line the pin goes to, if we are lucky its hiting another small resistor before it dissapears.
im not sure ifit works though, so better let somebod with modding skills do it :D
I have been saying this forever :)Quote:
Originally Posted by [_Silence_]
In my experiences....these QAPH ES cpu's can only scale with 17x. The RETAIL dothan chips can scale to 3900mhz with any multi you like. The ES's get unstable above 3250-3400mhz with any multi under 17x.
To pull serious power out of these ES cpu's, you need 3900mhz+ with 17x.
I was wondering why you would need a vcore mod? Isn't 1.7V + 0.08 overvolting enough?Quote:
Originally Posted by Neova
Thanks for your post k|ingp|n :toast:Quote:
Originally Posted by k|ngp|n
I was blaming the board for this problem and it seems to be a cpu problem...
Scaling with multi 17 the fsb will always be to low.. :(
Lets see max we can get from this chip with DICE...
No problem at all ojdr. It's definitely a chip issue for sure.Quote:
Originally Posted by ojdr2001
Good luck with your DI runs friend :toast:
200MHz boot only applies to i865's and i875's. Not the 915's.Quote:
Originally Posted by Always
hmmm thx for the heads up vince... thats a very weird issue though... hmmm weird... is the same issue present on the 865/875 boards with ct479? or only on the 915 chipset? hmmm
I've been having a similar problem with a 770 and the multipliers. Say a go for 11x, the cpu/mobo will go up to ~265MHz FSB giving just over 2.9GHz. Tho when I then move to 12x the CPU seems to freeze up as I approach 2.8GHz. The RAM has headroom but the chip won't go the same distance for some reason. Why is this? Has anyone else had problems here like this? Doesn't matter if I use a modded adaptor or not (have both :D)
thats very odd... this multiplier thing is confusing....
My original ES wont scale with any multi other 17x pat 3300mhz or so on any mobo. I have another sitting here that I havent tried yet....it will be interseting to see if that one is the same. Maybe I will get to try it someday ;)Quote:
Originally Posted by saaya
The ES chips that all the japanese overclockers were setting records with showed same behavior. YOu can see in all their 4ghz benches were only run with 17x.
FIRST LOOK of the P4GD1 mobo.......:)
When I set it at 120MHz PCI-E into bios, I can do up to 259MHz into windows......No more.....When I set it at 121MHz and above, I get no post.......2nd. When I boot at 200MHz and above, I have a Ram devider involved so 200MHz CPU and 160MHz Ram.......How to avoid it?........3rd. When I boot from 133MHz to 199MHz with auto for the rams, I have the 1:1 devider.......4th. It's almost identical with the ASUS P4P800-SE Mobo as for the performance BUT it can't "touch" the P4C800E-Deluxe.....5th. Rams does higher frequency than the P4P800 mobo....My "second pair" of the 256MB BH-6 did 259MHz easy with only 3.4Vdimm......They can do more with these low Vdimm but I got freezings above that.......and finaly 6th. I think I'm able to solve the high fsb freezing and not locking the PCI-E frequency by a HARDWARE Mod......;)