ocing a 1.33 axia

here is from another forum:

I finally got my axia 1.33 up to 1.5 but it crashes in tribes 2. I don't know why. I'm assuming there's perhaps a voltage setting that needs to be tinkered with? I'm on an asus a7m266 with 256 megs of pc2100 266 ddr ram. Maybe the mem can't handle the increase speed? I'm lost.
--------------------------------------------------------------------------------

increce the vcore(or in laymas terms the cpu volatage) to at least 1.750volts if you havent alredy.
you might also try incresing the i/o volatage to 3.4v

-----

Now, the guy above mentions increasing the i/o voltage to 3.4v but I'm not sure if that the VIO or VIO1 voltage. My VIO1 will only go up to 2.9 and my VIO will only go incrementally from 3.30 to 3.56 to 3.45. I'm not even sure if these are the voltages he's referrinf to but they appear to be. What's more my cpu voltage won't go to 1.750 and all voltage options seem to be listed in pairs such as 1.85/1.825. The closest I can guess is that I need to go with the jumper setting that would yield the voltage pair of 1.75/1.725. Incidentally I'm using an Asus a7m266. Apparently these two options correspond to certain cpu types but the manula doesn't specify which voltage goes with which chip. I'm using an axia 1.33 tbird and trying to get it to 1.5 stable.

What are your temperatures under load?

51-52C

Your default voltage should be 1.75. If you are aircooled, your temps are not real outrageous, but if you get the voltage up to 1.85, which you will probably need to hit 1.5 stable, heat is going to be a problem. I got my 1.33 up to 1507(11x137) but had to back off because of the hot weather we've been having until I put my air conditioner in. The heat brought ambient and watercooling solution temps way too high for stability.

I noticed you mentioned a multiplier. My intentions have been to avoid joining the l1 bridges simply because it seems to be too much of a pain in the ***. Right now I'm just trying to get the chip to 1.5 with fsb manipulation. My only question was about the dual voltages listed in the mb manual. for example, on my mb i have the voltage setting at 1.75 / 1.725. I'm not sure which is correct. Someone also mentioned io voltages which i'm unfamiliar with as my mb has two places for manipulating VIO1 and VIO.

Hey, I have the same problem! Or HAD.

I had my 1.3 @ 1.5 using 1.77 volts and Prime95 would run for ~1-2 days w/o crashing. I could have gone longer but I wanted to use my PC.

So I boot up tribes 2 and first thing it does is lock up. I just uped my voltage to my CPU and it worked. But temps rose up a bit.

FNG (May 08, 2001 11:24 p.m.):
I noticed you mentioned a multiplier. My intentions have been to avoid joining the l1 bridges simply because it seems to be too much of a pain in the ***. Right now I'm just trying to get the chip to 1.5 with fsb manipulation. My only question was about the dual voltages listed in the mb manual. for example, on my mb i have the voltage setting at 1.75 / 1.725. I'm not sure which is correct. Someone also mentioned io voltages which i'm unfamiliar with as my mb has two places for manipulating VIO1 and VIO.

Click to expand...

I am assuming that you don't, or don't want to unlock your multiplier. In all honesty, at least as far as my experience with the 1.33, going strictly with just FSB increases, you are going to have to get severe with voltage. I am giving you my experience with the chip only, because I don't have the same mobo that you do. if you have an AXIA like your post indicates, you have to find the "sweet spot" between multiplier and FSB increases. Your chip should already be unlocked. I am not stating Gospel, but every AXIA over 1G I have heard of is multiplier unlocked. Give it a try, and you might be pleasantly surprised.

well the bios definitely won't let me manipulate the multipliers and apparently to get them unlocked i need an electrically conductive pen, a mechanical pencil or the trace tape. i'm not quite up for this challenge just yet. but right now i'm at 1.4 and 3dmark will run. I have my voltage maxed out at 1.85 but when i try to run 3dmark at 1.5ghz it crashes to the desktop. The machine dones't crap out, just the benchmark. Prime95 will run at that speed for a bit, but that's boring. I'm trying to get it stable but the OS seems fine, its just that some software won't work at the higher speeds. I *guess* it's a voltage problem but I can't take it any higher. Maybe this is the part where I'm just supposed to connect the L1 bridges and go to town...

Prime95 will kick you out if your overall speed it too great for the voltage and temp you are using. It does not test the video very much, or sound card. Increasing the FSB will impact stability of the video and audio (PCI) bus as well as the cpu and memory. Playing some intense 3D game "covers the bases" missed by Prime95. If you have the FSB OC'd too much for your existing PCI/AGP/ISA devices, Prime95 may not shake that out. That's why it's good to have a "One-Two" punch for stability testing. I run Prime95 torture and then 3D Mark 2001. If you can get through both of them for as many passes as it takes your CPU temperature to stop rising, chances are you can rest assured that your OC settings are stable.

Just say MHO

Hoot

See this is where the problem comes in. I installed the latest AGP drivers from AMD but I still can't get through 3DMark. I can get my fsb up to 142 but when I hit 143 then 3DMark at some point just drops to the desktop. I have my voltage maxed and have even resorted to upping the clock generator voltage setting and the +2.5 voltage setting (VIO & VIO1 respectively). Still no dice. What you say makes sense but I can't think of a way around it. I've seen reports of other people with my mobo taking a 1ghz tbird up to 1.6. I can't even get my 1.33 past 1.42. Could TOO MUCH voltage be the problem? Perhaps if I take it back down to 1.75 on the vcore? I'm at a loss. I guess as a last resort I could reconnect the L1 bridges. On your signature you have a second voltage listed (1.85/3.57V) is that 3.57 the VIO voltage? aka Clock generator voltage setting? Thanks for all the advice.

BTW, is there a difference between just running Prime95 and a Prime95 torture test?

Ponder this similar analogy:
I have read of some people who can get their F-150 pickup truck to go 150 Mph, but mine only does 143 Mph. My F-150 must be defective.
Though excellent overclockers, the 1.0G/266 tbird has a certain degree of variability at the extreme top of its performance limit. Ditto on identical models of motherboard, memory, etc. At some point you have to sit back and say "Job well done". The degree of overclock you are achieving is applaudable. Be happy.

Hoot


Hoot

Hoot (May 09, 2001 12:02 p.m.):
Ponder this similar analogy:
-snip-
Though excellent overclockers, the 1.0G/266 tbird has a certain degree of variability at the extreme top of its performance limit.

Click to expand...

Oops, should have read 1.33G/266

Hoot

Thanks for the help and advice! but one last question... what is that second voltage listed ?? the 3.57V???

First off, that is the I/O voltage and it is what you adjust (slowly) upward, if you have an FSB related problem. If you can't solve the instability by the time you get to 3.8, chances are it is not the cause. Up to 140 Mhz FSB, my KT7A is happy at 3.47. 145 and 150 needs 3.57. The actual increments are 3.4, 3.5, 3.6... but my actual voltages are .07 higher than what is called for in bios, that's why I mention them.

Secondly, I was wrong in my first post. I had 1.0g/266 on my mind. A 1.33g/266 should be able to do 1500 without a lot of struggle. I am not familiar with PC2100 technology, so I don't knop how it compares to SDRAM above 133 Mhz in term of behavior. On the KT7A, we don't have the luxury of being able to adjust the 2.5V (Vi/o1), so I have no experience with how it impacts performance. Sorry I'm not much help on this. Definitely run the V i/o up to 3.5 and see if that stabilizes you in Tribes. Don't vary too many settings at once, or you won't be able to determine which one was the critical setting. I actually take a spiral notebook and keep notes on what settings yield what results. That way I don't have to "re-invent the wheel" every time I try something new. Also, it makes me feel like a scientist ;D
"Hey Moe, I'm a scientist! Bwoo bwoo bwoo!"

Hoot

haha. yes it's becoming quite a scientific process here as well. I know I should be able to hit 1.5 stable but I can't even get it to 1.43 stable as it is. I don't know if it's the voltage or what... my vio1 defaults to 2.7 but i kicked it up to 2.9. seems i was more stable at the intermediate 2.8 on the vio1. as far as vio i've got it maxed at 3.56 (it won't go to 3.8 so i don't know what to do there since you say up at 3.8 it should be stable if it's fsb). Guess I'll just keep trying diff combos but as far as everything else goes i'm at a loss.... thanks again for all the advice