made it to 2.5ghz on my barton

[subzero813]

just an update. this was my setup:

Athlon XP 3200+ Barton (x11 multiplier, 2497-2502MHz)
Asus A7N8X-E Deluxe @ 227mhz FSB
1GB PC3700 ram (gigaram mach 1000) @ 2-3-2-5 and 1:1
AGP @ 66mhz
vantec tornado on cpu, and another giant intake fan, plus throttle controller and everything's quiet on low rpm and still pushing plenty of air to keep it cool. also put a 40mm fan that blows 8-9cfm of air (using electrical tape, is this safe? it hasn't fallen off, looks stable...) on top of the Northbridge Heatsink.

had to put cpu voltage to 1.85 (1.80 in bios). ram was at 2.8v. this board doesn't let you change the motherboard voltage.

it wasn't stable. i knew it wouldn't be, i didn't bother running prime95. temperatures were good, so i just tried to get off a 3DMark05 run as fast as possible. it crashed to the desktop as soon as it hit the cpu tests, so i re-did it (took another 2 tries to finish the entire run, no BSOD though, so it was just getting errors in cpu calculations i guess).

i did also have my Gigabyte Radeon X800 XT overclocked to 520/560 (XT PE speeds) during this, with the fan @ 100%, temperatures never got over 63-64 C according to ATiTool during this run. normally i keep it at 500/500 though. because ATITool reports some crazy artifacting, even though i see nothing in 3DMark or other games, nor any high temps in ATITool's reading.

6088 3DMark05. blah... not that great of a jump. however...going from stock 2.2ghz to 2.35ghz gave me a great score increase all around, and a 4000+cpu mark score, and was rock solid. so i'm now at: 216MHz FSB, and cpu at 2377MHz, and tightened ram timings to 2-2-2-5 without any hiccups, problems, etc. Only ran Prime95 for like half an hour but...i figure that's stable enough. won't crash any of my games. temperatures are great. voltage is at 1.75 (1.70 in bios). i'll see how it sticks...it was actually quite stable at 218fsb and 2.4GHz even, but i wanted to be safer with the voltage.

throughout, i'd periodically get a BSOD, then i'd just up the cpu voltage another 0.5 in bios and it'd get past that, until i hit 2.5 then i stopped.

 

[Fresh Daemon]

Nice overclock - and quiet is always a bonus. Perhaps consider (if you haven't) some AS5 thermal grease, a Thermalright heatsink, and a Zalman passive northbridge cooler. They cool better than the dinky stock ones with the buzzy fan anyway, and miles better if you have a case fan blowing in their general direction. I don't know if you want to sink more money into your rig, but for a better overclock, that's what I'd do. You can probably find it used now that everyone's moving to A64.

 

[subzero813]

Originally posted by: Fresh Daemon
Nice overclock - and quiet is always a bonus. Perhaps consider (if you haven't) some AS5 thermal grease, a Thermalright heatsink, and a Zalman passive northbridge cooler. They cool better than the dinky stock ones with the buzzy fan anyway, and miles better if you have a case fan blowing in their general direction. I don't know if you want to sink more money into your rig, but for a better overclock, that's what I'd do. You can probably find it used now that everyone's moving to A64.

already got a swiftech mcx462v with AS5...i think its motherboard voltages that's holding me back.

but i can't explain the difference in cpumarks in 3DMark05...at 2.36 i had 4000+ barely, then it dropped to 3700 at 2.4ghz...because there's 2 tests. one got better, the other got worse....

 

[Fresh Daemon]

What changed in the rest of your system? Memory timings, FSB speed - anything? I would expect that if one test actually became slower at a higher CPU speed then something else has gotten slowed down enough to undo the benefit of the higher CPU clock. Look carefully at the speeds and timings.

 

[subzero813]

true, i went back to stock video card speeds (500/500) whereas that test was at 520/540. i guess the video card helped boost the cpu test by half a frame a second on the first test (the balloon one). the other test is usually same fps everytime.