ocing doesn't boost speed at all!

Quote:

---

Originally Posted by Holst

Nice one binkgle, your getting there now.

Try a little more voltage on the CPU, 1.75 is about as high as I would go. Just to see if you get get the system stable.

If that dosent help try lower memory timings (keep 1T, but try 2.5,4,4,10.1t)

If thats still no use try lowering the HTT a bit, try 295..

---

i bet this is a prob with the cpu, not the mem, as the mem is running @ 200 mhz (stock) when cpu is @ 300x8. i'll run memtest86+, though, just to see.

i'm gonna test it with prime95, too. i think i'd call it truly stable if it can do about an hour, maybe even an hour and a half. what do you think?

EDIT: 300x8 lasted 3 minutes in prime95, and 295x8 was reaching the same, very high, load temp (61-63 celcius), so i didn't wait to see how long that would last. i think for this i'm gonna need water cooling. i'd want to cool cpu and gpu, and, if possible, northbridge, and would prefer to start off with a kit rather than lots of seperate parts that i buy. what do you guys think of the coolance exos or the zalman reseator? for both i'd get a better gpu block. also, how would i make sure that it wouldn't leak?

Hmm, getting better but i've seen 3000+ NC's hit 2.7ghz on 1.6v. I would'nt doubt if your ram is causing some problems also. Did you try using 2.9v on the ram?

Your HTT speed dosen't increase perfromance what-so-ever as i've stated allready. HTT is the speed at which the CPU talkes to its componets, and the HTT speed is allready talking faster than the componets can, thus not increasing performance.

should i run ram at 2.89v when cpu is doing 300x8, cause ram is running at stock 200mhz? i'm sure my cooling has something to do with it.

Its not going to hurt it, I'd try less cpu mhz and more memory mhz.

Quote:

---

Originally Posted by WesM63

Its not going to hurt it, I'd try less cpu mhz and more memory mhz.

---

could you recommend a specific htt x multi and htt:ram divider? i'm getting lost in all this testing.

Quote:

---

Originally Posted by WesM63

Hmm, getting better but i've seen 3000+ NC's hit 2.7ghz on 1.6v. I would'nt doubt if your ram is causing some problems also. Did you try using 2.9v on the ram?

Your HTT speed dosen't increase perfromance what-so-ever as i've stated allready. HTT is the speed at which the CPU talkes to its componets, and the HTT speed is allready talking faster than the componets can, thus not increasing performance.

---

So your saying that 9x300HTT 3x, 133mhz divider at 2700mhz will not increase perf. compared to 13.5x200HTT 1:1 ratio????

Quote:

---

Originally Posted by The Virus

So your saying that 9x300HTT 3x, 133mhz divider at 2700mhz will not increase perf. compared to 13.5x200HTT 1:1 ratio????

---

when he said htt he was talking about the htt multiplier, and meant that having it set at 2x rather than 4x really has no performance impact at all.

this thread died over the past few days.
question: would 300x7 provide a noticeable perf boost over 204x10? ram would be @ 200 2-3-3-10-1T (maybe/probably 2-3-2)

EDIT: wasn't stable/too hot (mid 50's celcius within a minute of starting prime95 blend torture test) when running 300x7 1.75 volts, ram @ 200 mhz. i guess i just can't oc very much (i.e. not at all) effectively. i'll try again when i get hl2. if you've any more suggestions, please, please, please post them here.

2nd edit: would taking the fan controller of my hsf and letting it run at 2500 rpm as oppossed to 1500 help? would there be a noticeable cooling perf difference? noise level?

shoul be able run 300x7 at stock volts easily??

as long as CPU and ram speed are constant performance should be within a couple %.

i've tried running superpi with everything at stock and i gets errors when performing the 32M digit test. why? cpu @ 204x10 1.50v, ram @ 204 mhz 2-3-3-10-1T 2.75 volts.
i'll try 300x7 again

300x7 errored while running superpi 1M digit test, and about 30 sec in prime95. vcore was 1.65. ram was at 200 mhz, 2-3-3-10-1T, 2.75v

Try cas 2.5 or 3, maybe 2T, volts up or down... if its not stable you have to turn something down.

how much of a perf dif is there between cas2 and 2.5? 1T and 2T? why is 1T so much faster?

anyway, it wasn't a mem error, mem was at stock. i'm sure that was a cpu limitation.

cas latency doesnt matter too much, about as much as cpu multiplier.

2T takes a big hit.

1T is faster because it sends commands once per clock cycle, 2T is once every other clock cycle.

ive gotten up to 2.55ghz recently because of my new OCZvx i love it. i got the same proc as you. your ram clock & timeings suxxors.

i seem to be stable now with cpu @ 240x10 1.62v, ram @ 200mhz 2-3-3-10-1T 2.8v.

i hate my ram, but it was only $100, which fit my imit of $800 when i recently overhauled everything about my computer except drives, even case (well, hd died last week, so I bought a raptor :) ) my ram is only one step up from value (it's mushkin blue line, rated for 2-3-3 @ 200 mhz), so it's my major ocing bottleneck. i won't be able to get more ram for a while, unless i get a job and stop leeching from my bank account (which my mumhas control over, so i can't do anything without clearing it with her). where did you get your ocz vx ram? do you use their ram booster thing? what speed and timings do you run at? where did you get it? would boosting its voltage help its oc? it cannot, and this has been proven with this type of ram, be run over 204 mhz with 1T stably, which sucks.

i thought cpu multi was important (200x10 is much faster than 200x9, isn't it?)!

cpu multi is important.

10x200 with 200mhz mem is faster than 9x222 with 183mhz mem divider(9/10, 200mhz mem)

what exactly is cpu multi? how does it work?

i'm gonna try to squeeze proc up to 244x10, therefore ram to 204 mhz, and 2-3-2-10-1T (which are the timings i use at stock).

okay, 240x10 and 235x10 (mem at 200 and 195 respectively) cannt pass even a minute of prime95. what is this? mem is at stock speeds and timings, with extra voltage! cpu isn't overheating. it loads in mid to high 40's (celcius). why isn't this working?

Because your CPU isn't stable at those speeds.

any way to improve the situation?

There may be, there may not be, but this isn't the thread for it. Read some more overclocking guides.

i'm away until monday, but if anyone has any ideas, please post here!

270x9 1.62v seemed ok. mem @ 221 3-3-3-10-2T or 2.5-3-3-10-2T, 2.79v (though i might pump in an extra .1v). sandra cpu arithmetic was 11062, 3969/5157, mem bandwidth benchmark was 3008, 3013 when ram was @ 3-3-3-10. when i lowered ram to 2.5-3-3-10, mem benchie went up to 3025, 3026, but cpu went down to 10401, 3970/5160.

i'll test further for stability (so far i've only done sandra and a 1M calculation with super pi), but which of these ram timing setings means better perf?

EDIT: as a reference, with cpu @ 204x10 1.52v, mem @ 204 2-3-3-10-1T 2.74v, sandra cu arithmetic was 8779, 3332/4329, mem bandwidth was 3137, 3137

anyone?

oh yeah, and both of the below 270xp setups failed prime95 in under a minute