New to overclocking.. XP2500/ A78NX

[Ankun]

yes well his is rev1.04, so he must of flashed to get 1007. I'm just sayin he should try it at 1.04 if he hasnt already.

[mlc]

thanks guys...

I was running 1007 when I initially wrote in.. then for my second 2nd day of results I flashed to the 1004 Uber version. I've read, the 1004 has been the most consistently stable bios for overclocking when trying for fsb=200, particularly with the rev 1.04 boards..

I did manage to get it to boot up Windows with FSB=200 after going to 1004, and then gradually raising vcore to 1.85v and the mem voltage to 2.8... Sandra tests ran fine.. as did Prime95 benchmarking.. but I did receive errors during the very early stages of stress testing... Temps never exceeded 46c for diode and 44c for cpu socket...

not sure at this stage what makes sense to try...the memory timings were at 7-3-3-2.5T

[bblqj78]

I use IGB of Corsair XMS DDR400 and running at 11,3,2,2. Run you memory at CAS 2 for definite. Then try 11,3,2,2....as I manage above....and then try to go for 11,2,2,2.
You can try Sandra ememory benchmarks...but I found the best results for me at 11,3,2,2....my memory just didnt want to go to 11,2,2,2 regardless of voltage. You are using the Low Latency Stuff which is probbly better for getting lower timings. You shoudl at least manage the 11,3,3,2 a stock voltage....I am at 11,3,2,2 using stock voltage.

[LostBok]

FYI I'm running an ASUS A7N8x-E-Deluxe with a Barton XP2500+ @ 10 x 200Mhz, 1 x 512Mb Corsair XMS PC3500c2 (DDR433) @ 400Mhz, 2.0-3-3-6. All stock voltages. <40C unloaded and up to 44C loaded, ASUS Q-Fan at 4400rpm unloaded, not much quicker loaded, but haven't checked.

Waiting for delivery of 2 x twinX 512Mb PC4000c2 chips to arrive before I really bother trying to OC / tweak them, no point in putting myself through grief for the sake of <1 week's slightly faster performance.

LostBok

[mlc]

can't do the 10x200 due to locked multiplier... I've tried dropping it down from its 11.0x setting to 10, 9.5, etc.. with no luck... I assume you have an older xp2500 (i.e. prior to week 39)... which allows you to get those settings...

I've pretty much tried all of the suggestions I've received... the best I've been able to achieve is 11x200 with vcore at 1.85v and vdimm at 2.8.. However, while it does post... it fails/hangs within minutes during the prime95 torture test...

From a stability perspective, I seem to run fairly stable in the FSB = 188 to 192 range... but I'm still playing/tweaking at this point...

It is clear that running the older 1.04 rev board does suffer somewhat .... and that the uber 1004 bios ... along with the 1006 version as well, are best suited for overclocking the 1.04 version...

[Strawbs]

A7N8X Tutorials! Some will have already seen this, but for those that haven't; enjoy!

[Happy Joe]

Thanx Strawbs, alot of good info there.
Enjoy!

[mlc]

Strawbs...

yeah those are great.. I knew nothing about overclocking going into this thing... and basically did a crash course reading through all of those...

They do have something on the "wire trick" that supposedly lets you "unlock" some of the lower multipliers... but I don't think they apply in my situation...

[Happy Joe]

The wire trick is used to unlock the older AMD chips, it will not work on the locked Bartons, they use a different system of locking the multipliers.
Enjoy!