I been trying to get 1T stable for quiet awhile now.
I found the secret this morning by accident really.
without wasting your time...
NB to 333 with bios 0404.
use the 5:6 or 5:8 dividers.
have good ram. lots of patience, and cooling.
push the voltage hard.
tweak.
you know the routine.
Heres some 10K Bandwidth at only 3.1 Ghz with a Allendale.
1T is about 3-4 hundred more in bandwidth.
Nice... :).
Stable for SPI or 3DMark?
Wow, that is sick.
I am done benchmarking for the day.Quote:
Originally Posted by sobih25
all I focused on was breaking 10K.
maybe tommorow I will try, but tonight, I am just gonna relax, eat something good and be happy with such a good afternoon of tweaking.
these results on a scale of 1 to 10 for how hard they would be for anyone here to replicate, a 4.
so literally, anyone with a P5K Deluxe, bios 0404, some good D9 and some trial and error, could easily do this too.
How much vdimm did it take? I missed the whole 1T era so Ive no clue on how fast it really is.....
2.55. I was pushing everything the board could give me, without it being modded.
I really wish I had more voltage for it... I think this board can go much higher at 1T with more voltage.
You sure you are at 1T?
your at MUCH higher MHz... and MHz directly effects the ram bandwidth.
do that same test, at 3 ghz... with a 2 meg cache.
plus, look at the command rate.
you cannot change the command rate, within windows.
try it. you cant.
I was thinking you might be using the Maximizer, thats very cool for the boards stock 2.55v limit. Must try this out, thanks kunaak.
I would use the maximizer, but till I see someone else using it, without killing the board, or needing to mod it... I wont be using it.
I just dont know if the maximizer works on the P5K yet, and I dont have the cash to be the guini pig.
(how do you spell that anyways?)
guinea I think
My bad, dropped around 400 points at the same speed. I could see the 2M chip doing worse as well.
This board doesnt have a 1T option yet but it would be nice to compare or try some fat body 3-2-2-x 1T. I have some sick Muskin fat bodies
Yup spelt guinea :)
Yeah cpu mhz will effect everest bandwidth numbers here's 8x493FSB 1:1 @493mhz 4-4-4-9
http://fileshosts.com/intel/Asus/P5K..._bandwidth.png
But to achieve 10k bandwidth with just 3.1ghz is very nice :)
Guess there is some use to 0404 bios afterall :)
eva. come on,
do the 1T...
you know you want to.
Mobo: P5K Deluxe
Bios: 0311 (official)
Memory: OCZ HPC Edition 1gb x 2 kit
Voltage: 2.5vdimm.
1T require more vdimm as compared to 2T on same memory settings
http://img180.imageshack.us/img180/6...ndrajn4oc6.jpg
Nice ..but expect to see better bandwith - here is some 1T numbers 450Mhz 3-3-3-3..1T (1:1 of course) with 680i to compare
DFI 680i UT
X6800 Air Tuniq
Corsair 6400C3 2.5v
http://pedrorocha.planetaclix.pt/DFI...60_Everest.jpg
Superpi 1MB run
http://pedrorocha.planetaclix.pt/DFI...1_1_333_1T.jpg
;)
As Kunaak said earlier...try that at 3.1GHz with a 2MB L2 cache chip now. :p:
That's really something...1T at such speeds. memory's 600MHz shot is very impressive too. P35 is looking more interesting by the day.
Yes it is ... one reason I'm liking ddr3 :D :Quote:
Originally Posted by eva2000
http://s90.photobucket.com/albums/k2...0C6Everest.jpg
crazy...Quote:
Originally Posted by Aivas47a
my little 1080 mhz ram is nearly as fast as your 1440 mhz DDR3.
well... theoretically....
who knows about real life...
need someone with DDR2 and DDR3 + the P5K deluxe, and P5K3 Deluxe + lots of time on thier hands, and willing to test how 1T effects DDR2 and how it compares against DDR3...
if only someone at Asus felt like giving me some free hardware.
Yeah I'd really like to see some comparison tests like that from you or one of the other expert memory tweakers around here.
Just Curious...
but can anyone else do this...
3.1 ghz, and 10K bandwidth??
perferably with a 2 meg cache, since the 4 meg cache has a instant advantage with bandwidth too.
I know this was hard as hell for me to do...
I'd like to see how hard this was, if someone else might try it.
http://www.speedtime.cn/bbs/attachme...5oNBrM11Ph.jpg
http://www.speedtime.cn/bbs/attachme...VjZkY2Rv42.jpg
http://www.speedtime.cn/bbs/attachme...WmUcU26yoJ.jpg
Finally it can run DDR2-1170 4 4 4 4 1T but can not run any bench.
I dont know how much 4mb L2 cache makes a difference so I ran this at only 3000mhz.... I think at the end high mhz and tight timings still the best setting.... First try ... Rams are Ballistix Pc2 5300....Quote:
Originally Posted by Kunaak
http://ocidb.com/albums/userpics/100..._300000___.PNG
From an earlier test...@3600mhz...
http://ocidb.com/albums/userpics/10062/normal_11k.JPG
Settings are 32M stable :D
http://ocidb.com/albums/userpics/100...al_xx12_xx.PNG
http://img519.imageshack.us/img519/5...1000iy6.th.png http://img504.imageshack.us/img504/7...9571qn9.th.png
kentsfield + striker extreme + 2x1gb crucial 10th ae. difference between 1t/2t.
cannot use 1t on p5kdlx and these sticks :|.
Man when I attempted 1T on my P5KD I couldn't get anywhere.
For 1T performance my Striker did a lot better.
Maybe I'll give another go for 1T.