Unstable system...you know the deal, help!

[Placid]

I would guess the memory is sub-par or is having a compatability problem but I can't say that for sure the memory is the cause of your problems.
I would ask others with your board for their opinon.

 

[GooGobbler]

Here's something to chew on Placid:

I set the multiplier to 10.5 and FSB to 133 which is 1400 MHz and I set the memory to the most aggressive timings. RESULT: STABLE!

But when I set the multiplier to 8.5 and the FSB to 166 even with the memory at the slowest timings it crashes!

 

[Placid]

Strange stuff going on there.
Can you run 140 fsb stable,150?

 

[GooGobbler]

ill get back to you on that one....brb

 

[GooGobbler]

Yep I am running it at 150 FSB by 9.5 multiplier = 1425 MHz. It is stable runninng prime so far and mIRC doens't crash it when I right-click.

and memory = most aggressive

 

[Placid]

Sounds like the memory isn't liking running at 166.
Try 160 then 170 bump the mem voltage to 2.7v just skip 166.

 

[GooGobbler]

Alright, I tried 165/33 and it crashed. I tried 170/34 and it is stable! Oh and I didn't change the mem voltage, so it is still auto right now which is 2.500V I think...so...yea. Still want me to try 160?

 

[GooGobbler]

Alright, now I set it to 8.5*170 = 1445 and there is no stablility problems in prime so far.

 

[Placid]

No.
So it runs at 170 aggressive. But if you change it to anything other than aggressive it crashes even at lower speeds.
Sounds like the memory is ok.
Must be something odd going on with the bios or your board.
I guess you can live with it like it is and run 170 maybe higher and hope asus releases a bios version that will fix your odd problems or return the board.

 

[GooGobbler]

Yea, it doesn't seem to care for 166/33 or anything /33 unless it is the 1/4 divider which is 133/33. I am going to try increasing the multiplier now and see how it goes. Thanks Placid

 

[Placid]

Sounds like a good stick of 166 memory to run 170 at default voltage and aggressive timings.
Have fun

 

[GooGobbler]

Placid...big problem now. I tried to increase the FSB even more and I wasn't even able to go into my XP Profile. I had to decrease the FSB to 133 again and when I tried to log in it said my profile was corrupted! This is really ****ing me off now...... AKKAKAK!!!UGH

 

[Placid]

Stuff happens sometimes when overclocking
At least you have an idea of the max fsb you can run with aggressive timings.

 

[GooGobbler]

Placid, here is my final stable configuration thanks to your diagnosis

FSB: 170/34 MHz
Memory: 340 MHz
CPU Multiplier: 10.5x
Vcore: 1.775V
DRAM Voltage: 2.50V

170*10.5 = 1785 MHz

That's as far as I could get the multiplier up and increased voltage wouldn't help. Oh well, I think I'm satisfied.

 

[gulizi]

hi GooGobbler,
It is kind of weird that you can get stable at 170 but not 166.
I just set up a similar system same cpu XP1600 unlocked and the same memory as you, samsung DTL PC2700, my motherboard is MSI KT3-ultra ARU. DTL PC2700 is inferior to CTL which can do 200Mhz CAS2. but i heard you can still run it at 166 with agressive settings.
So here is my result : I tried to overclock to 166*10.5 at CAS2 agressive, I get crashes and won't pass prime95. I locate it to be memory problems cause I can't pass memtest-86. After some trial and error, I found if I set RAS to CAS delay to 3T instead of 2T, I can pass memtest-86. I would still get an error on Prime95 after a few hours. Still trying to figure it out whether it is the memory or CPU, ahh..