Pathetic O/C results winchester 3000+ on A8V

[entech]

Hiho,

I've tried to overclock the following:

Asus A8V
Winchester 3000+
2x512 Geil PC3200 (Rated 8-3-3-2.5)

and got really poor results. First i've tried to run 9x220 and mem in sync (with 2:1 ratio on A8V and 10-3-3-2.5 timings, 2.8 voltage). I've run Flight Simulator 2004 and of course it crashed after few minutes. So i've set the HT to 800mhz, pushed the FSB up to 240 and run the ram on 5:3 ratio so the mem will run at PC3200 level (i left the timings at 10-3-3-2.5). This time i've run FS2004 and it did run for 1,5 hours with no problems. so i left the computer overnight with Prime95 running the torture test (the maximum heat one). Unfortunately it crashed after an hour. I've dropped the FSB to 230, no improvement, prime95 crashed. Pushed the voltage to 1.425, 1.450, still no good. Upgraded the BIOS to 1.009, disabled AGP fast-writes, still nothing. The CPU is running as cool as 44C so the heat is rather not an issue, despite stock cooling. My power supply is Modecom 350W, nothing superb, but the voltages reported are stable also (the only weird thing is the 12v is reported as 11.49 - 11.51, but i've read that sometimes asus probe misreports the 12v and if it's stable with no significant fluctuations then it's OK). Do i just have a lame chip or is there anything else i can do to improve the o/c? Until now all i could achieve was 220mhz FSB with 5.3 ratio or 210FSB with 2:1 ... Prime95 runs fine on my chip with stock settings so it's not a problem with the program itself.

Thanx,

 

[superkdogg]

I can't really be much help for your scenario other than to tell you that people will say it's your PSU or your VIA chipset mobo. I'll refute either of those theories (although your 12v rail is light and could be the problem). I have a very similar setup with a 350W Fortron PSU with a "weak" 16A 12V rail, and an Abit A8V and have oc'd mine up to 272 FSB 2.45 (1.5 vCore) on stock air cooling, it actually has some more in it-but it runs 62-3 degrees during prime and I don't want to burn it up yet so I'm @ 266=2.4 until I get a better cooler.

Try lowering your multiplier and see what kind of an FSB number you can put up. Double check to make sure your AGP/PCI ratios are locked. If your fsb still maxes out @ around 230 when the mulitplier is lower, you have a memory or possible motherboard issue. If your FSB continues to climb but the chip speed hits the same ceiling, you may have a bum chip or your PSU could be letting you down.

 

[entech]

Well this is really strange. i can get to 266FSB with 3:2 divider and fly all the way from atlanta to orlando in FS2004, nothing crashes, this sim uses every resources possible at 100%. prime95 crashes after four loops or so like it used to. shall i believe in this 'p95 crashes on a64's urban legend? i've changed the PSU to friend's Huntkey 400W with 20A on the 12 rail and now I get 11,612 so slightly higher. On the previous PSU the system sometimes had problems coming up (you could press the PWR switch and nothing was happening), seems like this issue is gone too.see cpu-z snapshot

 

[superkdogg]

Did you write down the numbers from your heatspreader? Rumor has it that the earlier ones fail Prime for some reason as you noted. They are supposed to be stable otherwise although some dispute the stability of a CPU that can't pass prime although everything else may run well. Mine is a week 44 with no issues. My FSB is around yours, so it sounds pretty similar. Sounds like your RAM was holding you back before. Now you may be coming against heat. If prime errors out, you either have a CPU that's not stable at that speed, or you could be building up too much heat and causing some errors that don't trigger a reboot but that will crash prime. My own opinion would be that you're fine if none of the applications you use are crashing, but that I wouldn't make the mistake of calling your system "100% stable" if it causes prime crashes.

 

[entech]

well no i didnt't can i see it anywhere else maybe on the box or such? or maybe some software can read it? i don't really want to remove the heatsink, break the thermal compound so i 'd have to reapply it etc...

 

[superkdogg]

No big deal. The rest of my advice still stands.

 

[entech]

well the maximum observed temperature was 45C. is it overheating? also i'm running 1.425 voltage though the asusprobe reports 1.44 - 1.456 - 1.72 fluctuating. this is weird. should i put the voltage higher? won't it damage the chip with such fluctuations? what can cause them, the psu is now a rather good one so..

 

[superkdogg]

You shouldn't have that big of voltage fluxuations. A .3 volt difference between high and low is way too much. It might be that the PSU is not working the way it should, or your CPU or MOBO or, most likely the probe is a little wacky. Many software monitoring programs can misreport things. That being said, I wouldn't raise the voltage any more. 45C is definitely not overheating that CPU. 60-65C or so is the start of the red zone. The only other thing I can really tell you is to read as much as you can about OC'ing and ask specific questions when you hit a snag. There are stickies on the forums on basic guides which you might find educational and there are other places to check out too. Best of luck!

 

[Mikeyflan]

try and get a new PSU something better then what you got, look in to a antec true480 or something

 

[entech]

Hi,

I've tried to run it 240FSB x 8 multiplier and it crashed just as on x9. Is it the board issue? i can switch it for the MSI NEO2 for a little money, is it worth it? I read some people on various forums think MSI is crap and find the A8V much better... i'm confused now... I can also leave the asus in and switch the 3000+ for the 3200+...