![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
__hollywood|meow
[H]ard|Gawd
- Joined
- Feb 20, 2006
- Messages
- 1,500
hello all. i just received my KACAE 0605 APAW 3700+ e4 stepping san diego proc a couple days ago. these are supposed to be the best stepping chips for air overclocking, as they generally go very far on stock voltage and respond well up to around or below 1.5v.. but the lower, the better..
let me show you guys a screenshot of my experimentation so far.
stock 2.2ghz gave me 5500 in 3dmark05, and these overclocked settings you see gave me just over 6k with my new x800XT at stock (which by the way in preliminary overclocking results goes to about 540 core 525 mem with a nice load temp of ~60C). i think i'm doing pretty well at 2.7ghz 1.35v, 300HTT, ram at 245mhz 2.3.3.8 2.9v. my cpu tops out at 2750mhz or so and 330HTT (independently tested), stock volts.
anyway, my ram is TCC"X" (straight from ramguy on corsairs forums); definately holding me back. 245mhz at 2.8v gave DOS memtest errors, so i gave it 2.9 and it stabilized. would it be safe to give them 3.0 volts? i bet they would go farther if i did, but i don't know safe voltage tolerances for this TCC-whatever. i'm guessing 3.1 would be the farthest i should go.
so 245 did windows memtest eight hours with no errors at CAS 2, but CAS 2.5 at same volts and same latencies errored within ten minutes of the LESS strenuous DOS memtest. i really don't understand that part of it, but i've heard that a64s sometimes like one or the other, but not both.. is this true? am i seriously not going to be able to run CAS 2.5? are there timings to tweak to fix this? would a voltage increase help? i'd be very interested to know.
oh, and i realize my a64tweaker timings are loose in some places (async latency especially, but also idle cycle, row refresh, precharge, and recovery), but i've just been setting up to see how far i can go; afterwards, i'll work on tightening those timings.
i'll post back with more results later.. because next round is going to involve upping the CPU voltage to see if i can get to 3ghz >)
let me show you guys a screenshot of my experimentation so far.
stock 2.2ghz gave me 5500 in 3dmark05, and these overclocked settings you see gave me just over 6k with my new x800XT at stock (which by the way in preliminary overclocking results goes to about 540 core 525 mem with a nice load temp of ~60C). i think i'm doing pretty well at 2.7ghz 1.35v, 300HTT, ram at 245mhz 2.3.3.8 2.9v. my cpu tops out at 2750mhz or so and 330HTT (independently tested), stock volts.
anyway, my ram is TCC"X" (straight from ramguy on corsairs forums); definately holding me back. 245mhz at 2.8v gave DOS memtest errors, so i gave it 2.9 and it stabilized. would it be safe to give them 3.0 volts? i bet they would go farther if i did, but i don't know safe voltage tolerances for this TCC-whatever. i'm guessing 3.1 would be the farthest i should go.
so 245 did windows memtest eight hours with no errors at CAS 2, but CAS 2.5 at same volts and same latencies errored within ten minutes of the LESS strenuous DOS memtest. i really don't understand that part of it, but i've heard that a64s sometimes like one or the other, but not both.. is this true? am i seriously not going to be able to run CAS 2.5? are there timings to tweak to fix this? would a voltage increase help? i'd be very interested to know.
oh, and i realize my a64tweaker timings are loose in some places (async latency especially, but also idle cycle, row refresh, precharge, and recovery), but i've just been setting up to see how far i can go; afterwards, i'll work on tightening those timings.
i'll post back with more results later.. because next round is going to involve upping the CPU voltage to see if i can get to 3ghz >)