Final those Geil sticks can go very high. After test them with Hipro they also can go in my machin as high as well........
310fsb 1:1 with 7-4-4-2.5 and a very nice Sandra mem score..
7435...
Printable View
Final those Geil sticks can go very high. After test them with Hipro they also can go in my machin as high as well........
310fsb 1:1 with 7-4-4-2.5 and a very nice Sandra mem score..
7435...
And a screen............................
And a nice Unbuferd score also
Voltage?
Really good results reguardless though. Hipro kind of ruined the surprise you would have had though :p:
:) 3.45V but they have some more to go we will see.....................:)
Mmmm make me wonder if the life of BH5/6 has ended now. These new chips are (at last) providing higher bandwidth and higher 1:1 clocks with much lower voltage.
And all this just as the life of DDR(1) is about to end.
Oh 3.45v...maybe not but still higher bandwidth and 1:1
But remeber guys, those arnt Geil PC4200 sticks....
No... they are magical Hipro sticks!!!
Good job. lets see how far they can go!
What are they then?
His signature says PC4200
lol..... they are hipros 4200 ;)
But they are still GeIL PC4200 with CH5 right?
yes but its not ch5, its hynix or something
No they are just GEIL PC 4200 with the original heatspreaders and they are mine. With Hipro we just test them and they were able to make an 7500 Sandra mem score at 312fsb....:)
Very nice dude! :)
The unbuffered is a little bit low for 1:1 I think.
Good clocks :toast:
3500/3550 at 310 1:1 is nice?Quote:
And a nice Unbuferd score also
3700/3750 at 232 1:1 is a lot nicer IMO :D
thats teh beuty of tight timings.
but huge mhz are just as fun!
Ehm.. the pic shows 232 5:4, is that a bug / wrong mem detection by CPU-Z or the REAL unbuffered bandwith /w that 3.4C ?? :eek:Quote:
Originally posted by macci
Good clocks :toast:
3500/3550 at 310 1:1 is nice?
3700/3750 at 232 1:1 is a lot nicer IMO :D
Is 4100/4122 nice at 272 1:1 /w 2.4C ? :D
btw are these 4 sticks stable /w PAT and high FSBs ?
mine aren't really stable :rolleyes:
http://mitglied.lycos.de/coolinglimi...3/260_2048.PNG
/edit: 4 sticks screenshot
2.55v ...sick... just sick...
Bandwidht is real. the CPU-Z reports incorrect divider - its 232 1:1 in reality. Its on AI7 and so in order to run w/ FULL PAT it can't use 800FSB Strap and that is something CPU-Z can't read properly.Quote:
Ehm.. the pic shows 232 5:4, is that a bug / wrong mem detection by CPU-Z or the REAL unbuffered bandwith /w that 3.4C ??
I've only ran SuperPi 1M at 239.2FSB but that was the CPU limit. I have an ES but AI7 wont let me adjust multiplier so not much testing done w/ the board yet.Quote:
btw are these 4 sticks stable /w PAT and high FSBs ?
mine aren't really stable
ok thx!Quote:
Originally posted by macci
Bandwidht is real. the CPU-Z reports incorrect divider - its 232 1:1 in reality. Its on AI7 and so in order to run w/ FULL PAT it can't use 800FSB Strap and that is something CPU-Z can't read properly.
mmh, ok 239 is not SO high! ;) seems like the P4C800-E doesn't like 4 sticks, its stable until 250, at 255 Sandra crashed. With 2 Sticks I can bench /w nearly 20MHz more.Quote:
Originally posted by macci
I've only ran SuperPi 1M at 239.2FSB but that was the CPU limit. I have an ES but AI7 wont let me adjust multiplier so not much testing done w/ the board yet. [/B]
With 1 stick I can bench at ~282MHz, with 2 sticks (DC) I can bench at ~263MHz, and with 4 Sticks I can bench at ~250MHz.
I found out that this kind of loosing mhz is only with these damn Winbond Chips, other Rams is running with exactly the same speed in single and in dual channel mode.
Maybe its the i875P who is limiting?
:D yes but it's at 310. With that CPU i can't go so High but with another maby yes. About the Buferd score maby i can pass 7500:DQuote:
3500/3550 at 310 1:1 is nice?
Like I said 239 was the CPU limit (aircooled). It might go higher and it might not go higher hehe :D
w/ BH5 and 2-2-2 timings its quite often the NB/MCH thats limiting the max OC after certain point. For instance the AI7 maxes at around 225 w/ every setting put to max (F1-e-4-e-e) w/ 2x512 - 4x256 didn't even load OS w/ those settings.
Also my pair of KHX3200 does a max of 257 on IC7, 262 on P4C800 and 270 on P4C800-E.
The mobo RAM powersupply might be at the limits in some cases too: on P4C800 for instance the Vddr (under load - memtest86+) is at 3.1V w/ 1x256, at 3.05V w/ 2x256 and 3.0V w/ 2x512..
This is a nice unbuferd score with Mushkin PC3200 BH-6 at 254fsb 1:1 with 3.50V and Street Racer-Ench-5-En-En. Those sticks are not ready yet because they want some more burn in:D
hmm, nice fsb :)
Interesting, thx! :)Quote:
Originally posted by macci
Also my pair of KHX3200 does a max of 257 on IC7, 262 on P4C800 and 270 on P4C800-E.
I gained ~ 8 MHz Ram from switching P4P800 to P4C800-E. But I lost ~ 60MHz CPU when comparing IC7 to P4C800-E. I try to modding the 3-Phase VRMs.
Thats why I cool it down with some heatsinks! ;)Quote:
Originally posted by macci
The mobo RAM powersupply might be at the limits in some cases too: on P4C800 for instance the Vddr (under load - memtest86+) is at 3.1V w/ 1x256, at 3.05V w/ 2x256 and 3.0V w/ 2x512.. [/B]
http://mitglied.lycos.de/coolinglimi...of_Ram-MHz.jpg
I'm thinking about replacing the mobo Ram powersupplyparts with more strongly puppies! :D
Will this improving Ram OCing Macci ?
Nope...Quote:
Originally posted by andyddr
Mmmm make me wonder if the life of BH5/6 has ended now. These new chips are (at last) providing higher bandwidth and higher 1:1 clocks with much lower voltage.
Because if you took that same system running 310 mhz FSB and put in BH5 at the same voltage and ran it at 5:4 / 2-2-2-5 (248mhz) your system would perform better. :D
Everytime I think BH5/6 is dead I see another reason to buy some (macci). Jeez now I just don`t know what`s best anymore.
High FSB vs tight timings?
Yes you are right but the reason here is a top Sandra Buferd score and that score i think that it isn't bad...one.:DQuote:
Because if you took that same system running 310 mhz FSB and put in BH5 at the same voltage and ran it at 5:4 / 2-2-2-5 (248mhz) your system would perform better.
Certainly not bad at all :DQuote:
Originally posted by aristidis
Yes you are right but the reason here is a top Sandra Buferd score and that score i think that it isn't bad...one.:D
310 at 1:1 or 5:4 is always a respectable feat !
Nice job !
my bandwidth with a 2.4C and P4C800-DQuote:
Originally posted by andyddr
Mmmm make me wonder if the life of BH5/6 has ended now. These new chips are (at last) providing higher bandwidth and higher 1:1 clocks with much lower voltage.
And all this just as the life of DDR(1) is about to end.
http://membres.lycos.fr/cpulloverclo...emorybench.JPG
I had similar clocks to ya, but I never was able to bench Sandra at >316MHz, the Rig never posts beyound that and Sandra did not recognize the real FSB when adjusting the FSB in Win.Quote:
Originally posted by cpulloverclock
my bandwidth with a 2.4C and P4C800-D
http://membres.lycos.fr/cpulloverclo...emorybench.JPG
It was stable at 325/260! :)
http://mitglied.lycos.de/coolinglimi...max_stable.png
I wish I had a buffered Sandra bench here :confused:
This is something your grandmother can do: 283 fsb, 1:1 at 2.9 VDIMM with Vitesta PC4000:
This is 248, 1:1, 2-2-2-5 with XMS3500 at 3.2 VDIMM (4PDA2+ with mem tweaks for full PAT):
And a small update on Sandra 7500...:D
Mmm yeah I need new ram :|
dang those are some nice clocks
Why can´t we just switch... You get my Geil 4200 and i get yours..OK ? ;)
:D Ok this a good idea,...................but a long distance...:DQuote:
Why can´t we just switch... You get my Geil 4200 and i get yours..OK ?