New WR DDR636 2.5-3-3-5 1T PI 32M

Printable View

Show 100 post(s) from this thread on one page

10-24-2005, 07:43 AM
targ

Quote:

Originally Posted by LoKi2k

alright andre, i am going to get you today. my gskill 4400le can run 320 2-hours prime blend ;)

@2.5-3-3-5??

if so run SPI 32M and clam your WR.

looks like the TCCD race has started again for some reason...

i know gigi has some more to come.

targ
10-24-2005, 08:41 AM
AndreYang

Quote:

Originally Posted by kiwi

What about retail GEIL ONE 5xx? Can we expect at least 300 2.5-3-3-6 on average?

Maybe u need to handpick about 20 pairs from Retail GeIL One.
10-24-2005, 10:26 AM
gigi77

nice INDECENT SPEED MY Friends!

Good Luck for 320MHZ
10-24-2005, 12:53 PM
szukalski

Quote:

Originally Posted by AndreYang

Maybe u need to handpick about 20 pairs from Retail GeIL One.

Took me two pairs to find one that did 310MHz @ 2.5-3-3-6 @ 2.8v (300MHz @ 2.7v).
Could be that someone else already went through the first 18. :D
10-24-2005, 01:46 PM
s7e9h3n

I've said it before, and I'll say it again....no offense, but to compare TCCD clocks not only do you have to run the same speed on the memory, but use a mult @ 10+. Anything lower is much, much easier to do. Regardless, that's still quite an accomplishment Andre :toast:
10-24-2005, 01:48 PM
Vapor

Steve.....not all CPUs can do that. Isn't still an equal test of the memory using the 9x multi?
10-24-2005, 02:34 PM
targ

Quote:

Originally Posted by s7e9h3n

I've said it before, and I'll say it again....no offense, but to compare TCCD clocks not only do you have to run the same speed on the memory, but use a mult @ 10+. Anything lower is much, much easier to do. Regardless, that's still quite an accomplishment Andre :toast:

yes i dont quite understand this statement, the IMC shouldnt be affected by multi, and even if it was thats nothing to do with the ram itself being stable.

could you please explain more?

targ
10-24-2005, 02:42 PM
AndreYang

Quote:

Originally Posted by szukalski

Took me two pairs to find one that did 310MHz @ 2.5-3-3-6 @ 2.8v (300MHz @ 2.7v).
Could be that someone else already went through the first 18. :D

If i use bad A64 Tweaker Timing run , i will be running higher. Why am i done it?Because that's lose WR and test ram significance.
10-24-2005, 02:49 PM
s7e9h3n

Quote:

Originally Posted by Vapor

Steve.....not all CPUs can do that. Isn't still an equal test of the memory using the 9x multi?

NO.
10-24-2005, 02:52 PM
Vapor

Care to explain....I can completely understand the IMC getting stressed more with higher multis, but the memory too?
10-24-2005, 02:59 PM
bachus_anonym

It might not be as obvious as you think it is, Steve. It's not 100% guaranteed that If one does 315MHz on x10. then it will do 320Mhz on x8. It's not that easy.

I've been playing with my 4400LE for the last week or so. It does 32M @ 304Mhz 6-3-3-2.4-1t with x11 multi... So, far I used all tricks I know and cannot get it to pass 306Mhz, even on multi x7. I will still be trying but I 'm not 100% sure that what you claim is holy truth. If lowering multi gives advantage, then it's very slim...
But that's only my :2cents:
10-24-2005, 03:04 PM
TheMeatFrog

Quote:

Originally Posted by bachus_anonym

It might not be as obvious as you think it is, Steve. It's not 100% guaranteed that If one does 315MHz on x10. then it will do 320Mhz on x8. It's not that easy.

I've been playing with my 4400LE for the last week or so. It does 32M @ 304Mhz 6-3-3-2.4-1t with x11 multi... So, far I used all tricks I know and cannot get it to pass 306Mhz, even on multi x7. I will still be trying but I 'm not 100% sure that what you claim is holy truth. If lowering multi gives advantage, then it's very slim...
But that's only my :2cents:

Same here bachus, my FF do 307 2.5-3-3-6 regardless of multi :slap: 9x and 8x do nothing for me :confused:
10-24-2005, 03:08 PM
amrgb

From my experience if the limiting factor is the ram and not the memory controller, the higher the multiplier and the faster the cpu the better.

Thing is with my winnie at 2440 (8x) for 305HTT it took more than 30 minutes to complete a 32M run. When you are at the verge of stability (32M-wise) the faster you do the job the smaller the probability of getting an error.

I've tried with 7x and 6x and not only I didn't gain one single Mhz but also sometimes I could not even get a 32M run at 305.

How bad I wished I had a fast cpu at that time...
10-24-2005, 03:37 PM
Onepagebook

Quote:

Originally Posted by Vapor

Care to explain....I can completely understand the IMC getting stressed more with higher multis, but the memory too?

It doesn't mean memory getting stressed. It's all about the memory efficiency.
If Steve is wrong, then I think there is no reason we need run some Sandra memory bandwidth for own reference.;) From that bench tell us, lower multi. got lower memory controller efficiency. Consequently, under about the same FSB setting, using lower multiplier to bench is way not harder than higher multiplier does same FSB.
10-24-2005, 04:29 PM
Vapor

My argument was more clearly expressed by bachus....I was simply talking about stability, not performance. I can get nearly Sandra score @ 264x11 as I do 274x9 (actually, the former is higher), that I know and understand. Very similarly to TMF, my LA can do 306 2.5-3-3-0 as long as my CPU can clock that high....
10-25-2005, 12:58 AM
szukalski

Quote:

Originally Posted by AndreYang

If i use bad A64 Tweaker Timing run , i will be running higher. Why am i done it?Because that's lose WR and test ram significance.

? I don't follow here. :confused:
10-25-2005, 01:37 AM
hicookie

great job,dude, new WR!!
10-25-2005, 03:22 AM
kktham

Quote:

Originally Posted by AndreYang

519 TCCD

bro u pry the heatsink open to see the week of the chip??
10-25-2005, 03:41 AM
Vapor

As a matter of fact....his heatspreaders are off. There are pictures on the earlier page(s) (depends on view settings).
10-25-2005, 04:03 AM
gigi77

Quote:

Originally Posted by Onepagebook

It doesn't mean memory getting stressed. It's all about the memory efficiency.
If Steve is wrong, then I think there is no reason we need run some Sandra memory bandwidth for own reference.;) From that bench tell us, lower multi. got lower memory controller efficiency. Consequently, under about the same FSB setting, using lower multiplier to bench is way not harder than higher multiplier does same FSB.

go for 320 2.5-3-3-5

:)
10-25-2005, 05:07 AM
amrgb

:idea:
10-25-2005, 05:10 AM
Vapor

huh?
10-25-2005, 06:32 AM
AndreYang

Quote:

Originally Posted by szukalski

? I don't follow here. :confused:

So many people change A64tweaker after running 32M. If u care about 32M Speed. you will be found why their 32M are so slow.

Use this A64tweaker and run 32M again.

http://andreyang.myweb.hinet.net/A64.JPG
10-25-2005, 08:37 AM
ianocean

100Mhz 1.95us -what TREF is this displayed in DFI BIOS
10-25-2005, 08:48 AM
Onepagebook

Quote:

Originally Posted by gigi77

Hello Mr. Onepagebook,

:) :toast:

hi Master gigi77, your run and everything is more impressive :toast: , at least you won't get the shxty ram program them first(oh ya , pry the heatspreaderand from geil reinstall on the shxty ram) then return to store and cost you nothing. I should learn this huh? :toast:

Show 100 post(s) from this thread on one page

All times are GMT -8. The time now is 04:17 AM.

XtremeSystems