No unfortunately I've sold all cpus that can do 600fsb. I don't see any reason to bench s775 since sandy came :/ Vmch didn't seem to matter, kept it around 1.35. But dram ref voltage mattered a lot, sweet spot is -20mV on ch A, B.
Printable View
No unfortunately I've sold all cpus that can do 600fsb. I don't see any reason to bench s775 since sandy came :/ Vmch didn't seem to matter, kept it around 1.35. But dram ref voltage mattered a lot, sweet spot is -20mV on ch A, B.
Too bad! :(
40 euros
I got the reply
1250MHz and timings like DDR3
anyway this is what he called stability
http://img853.imageshack.us/img853/8...pturegd.th.jpg
Unfortunatelly i am not going to buy :D
DDR2 714 cas5 on AIR !!
714mhz C5-5-5-35 2T Just have a look hommies
2x1024mb CRUCIAL Ballistix Tracer DDR2 PC8500 D9 gmh (16F17) on GA-EP45-UD3P with 2,6v :yepp:
http://www.kirikoo.net/images/1EeKy%...503-230048.jpg
A little preview of the BW :cool:
Excellent bandwith, by chance I got a new board friday and had a 1:2 Test-run with some ballistix I found in the closet^^- are these your old Tracers? They are far better than you said :)
http://www.abload.de/thumb/ballistix...2m224vapeb.jpg
should go & dig up my old 700Mhz valid...
who woulda thought OCZ made some decent D9 sticks :ROTF:
Suicide overclocks did that chip degrade overall with that vcore? Balis are great to bad they never made 2 gb sticks and only 1gb per stick.
Yeah it's the firsts Tracers but what you mean with "far better than I said" ? for CAS4 ? I need to check seriously sorry...
Hey man the chip is under SS here, and it can bench easy at that speed ;) No suicide run in here, trust me ;)
Lol Sorry, you have a better memory than me :p I forget about this post last year.
No real improvement at last, the board isn't as good as the first one. Bw is high because of the CPU under SS and some tweaks on XP maybe...
So i've had a few kits of d9's for a while now, but never bothered to post here. Now i will.
Websmile seems to have the most amazing kits around, but i have a few that aren't too shabby.
EDIT: Updated with non-bugged divider, still 16fd5, 700mhz @2.54v for 32m, 2.4v for 1m.
Non-bugged divider run:
http://i929.photobucket.com/albums/a...zactual32m.jpg
Please check what I posted in the 700 MHz club thread! ;)
I dont know what is rated GOOD speeds on these D9GKX ic's, but mine is now doing pretty well running spi 32M with 1200MHz 5-5-4-10 with 2.4volts bios showing :)
Gotta try how far i can pull these with 2.4v 5-5-5-15. How much voltages is still safe on spi runs?
1000MHz 4-4-4-8 was ok too, didnt try more yet.
Sticks is Buffalo Firestix rated 800Mhz 4-4-4-12
At 2.40V you should be at least 50 MHz higher on both CLs.
For example my Crucial 10th Anniversary Kit that I use for benching needs 2.06V for 600 MHz 5-5-5-15 and 2.13V for 500 MHz 4-4-4-12, and it's just a mediocre kit, far from what I would call a good kit.
Can you try what your kit needs for these settings?
I've given my kit 2.65V for a short time while benching high CL4 on 1:1 (1M, MaxxMem), for 32M my personal limit is closer to ~2.50V though...
Don Dan: Yes i can try what i need for those, atleast tomorrow if i cant run it fast now. Have to go sleep soon.
I try that 500 CL4 now and other one tomorrow. Il post soon!
Edit:
2.12v 4-4-4-12 1000MHz Fail.
2.016v 5-5-5-15 1200MHz didnt even boot in windows.
il try more tomorrow. I have one other patriot 2x1gb gkx kit and 4x axerams too so more is coming :)
Obviously this kit isnt any special, but atleast this was cheap :D
SPI32m 682 MHz @ 5-5-5-15 @ 2.30v
PII 975 BE + DFI 790FX + 2x1 BallistiX 16FD5
http://www.abload.de/thumb/682gd889.png
Sorry for necroposting but I think I've something rather interesting to share with my fellow DDR2 enthusiasts here on xs. :D
E6750 - FSB 545x8 (not that good) single stage cooled
Gigabyte EP45-Extreme - BIOS F8
Cellshock PC8000 naked - D9-GKX - 1308 MHz 4-4-4-4 trfc 26 trrd 1 PL 7 Super-PI 32M stable
VDIMM 3 volt
http://uwot.eu/imagehosting/image.php?dm=GR9B
Looks like the board is also pretty fast: http://hwbot.org/submission/2577717_...min_1sec_968ms
Very good D9, but the voltage is hard :D - Cellshock used good PCB, looks as if your sticks can take it :)
Yes, they scale very well with voltage.
2.8 v -> crash at loop 3/4
2.9 v -> crash at loop 11
3.0 v -> pass
Too bad the board doesn't let me feed them with more juice. :D
Nice to see some good old DDR2 kickin ass Mafio :) Interesting to see such good performance from P45 too, thought only Biostar could manage such 32m efficiency. I tried Rampage formula with various good kits some days ago, but nothing close to this. I have some Dominator 1111 4-4-4 and cellshock 1000 4-4-4 sticks that should go really high.
Thanks, socket 775 is still a lot of fun.
The motherboard is also pretty fast with 45nm CPUs, here is a Q9550 at 520x8.5, Cellshock 624 MHz 4-4-4-4 and PL 7 32M stable.
http://hwbot.org/submission/2576098_
P.S. Can't believe it, yesterday's morning a CPU power phase completely blew up - both driver and mosfets - while benching the E6750 single stage cooled.
The motherboard shutdown in a strange way during the boot, at first I thought the CPU died (I was using VPLL 2.6 v :D ) but after a closer look I found out it was the motherboard...VCORE was only 1.7 v and I had a fan on the PWM zone...f(l)uck
For some kind of miracle both the CPU and the motherboard are still working so looks like now I have the only EP45-Extreme with 5 CPU phases. :D
I have another or two of those boards I can take off parts but I think I am missing the right tool (hot air station) to change the mosfets.