posted
200 2T 2.5-4-3-7 should go over 300, but I had much trouble on the 9NDA3+ rev2.x... 250-260 was the best I could do with several brands of 2X512...
-------------------- Too Many Computers,... Too Little Time ..... Com'on ???..! Posts: 27985 | From: Fire Island, NY | Registered: Feb 2003
posted
Must have gotten a wierd pair of xblk dimms. Makes no difference which slots or in which order they go in my rev2 9nda3+. Pass memtest at 305, 1:1, 2T, 2.5-4-3-6. And that's at 2.8vdimm. Board is unmodded, so can't try 3.0.
2T is about a 10% bandwidth hit and you always do 10% better in mem speed with it...
@ 45 Auto: Sounds like you pair is match perfectly! My kits alway did a little better swapping them around... 2 singles I got made a big difference swapping them around...
I must have a bad rev2... It is not stable past 250 with 2X512 of any brand, at any settings...
-------------------- Too Many Computers,... Too Little Time ..... Com'on ???..! Posts: 27985 | From: Fire Island, NY | Registered: Feb 2003
posted
Well...tried using slots 3+4, then 4+3 and the computer wouldnt even boot. Didnt bother trying to reset the CMOS though between attempts, but the CMOS reset automatically for the 3+4 attempt. I do believe 1+2 and 267 are gonna be my best for 2.5-3-3-8. Not too shabby though, cant complain much
posted
Yea...upgraded I sold the 9NDA3+ Ultra on Ebay and it was recognised by someone who asked if I was Kenny from the Rebels Haven. That made my day!
Kept it as a thumbnail to make loading easy. Plus the original image was massive..
IMHO the 267 "cap" is connected to the 533 rating. 267x2 is 534. Cant be a coincidence. I found that the idle cycle limit and refresh rate make no difference to synthetic bandwidth or superPi times, but increasing them let me push the speed up.