Just got an E6600 L630F049 & now a L632B173

[n7]

Installed the L632B173 this morning; added some preliminary results.

 

[Thor86]

Originally posted by: n7
Installed the L632B173 this morning; added some preliminary results.

Nice, looking forward to seeing your 3.6Ghz oc results.

 

[allies]

Originally posted by: n7
Ah don't mention financial things
I don't have the money i've wasted...

I'm an absolute fool for buying these two components when i could merrily run my E6300 + TT BT at extremely quiet levels just fine.

I blame the desire for more (which i completely do not need) on forums
Damn hardware addiction.

I know how you feel... there's always a thirst for more! I'm fine with my computer for now though... I think...

Nice work with the B stepping... 3.3GHz at 1.25V is sweet. I would love to have that, but I guess I have to "deal" with 3.0GHz at 1.425 V

 

[aimforsilence]

Hey Nate!! (N7) if you want more there will probably be a E6700 / Extreme edition with your name on it come intel price drop day

lol

 

[WoodButcher]

Originally posted by: n7
Ah don't mention financial things
I don't have the money i've wasted...

I'm an absolute fool for buying these two components when i could merrily run my E6300 + TT BT at extremely quiet levels just fine.

I blame the desire for more (which i completely do not need) on forums
Damn hardware addiction.

Hahahahaha!!! Join the club, We meet weds nights @ http%^&&**&%^$$###$%$%, It is coded because were another one of those anonymous groups. don't tell anyone!

 

[n7]

Holy crap i came home from drinks w/ teh buddies from werk, & Orthos hadn't failed :Q

10 hours worth of Orthos blend @ 1.25V
http://img296.imageshack.us/img296/8316/e6600b3303ghz9x36713vbike9.jpg

If anything, this B code E6600 is good for nice low voltage results.

 

[CloE]

nice B chip, I also brought 2 e6600. only can keep one, I have L630B and L632A. they are in UPS track now, dont know which one it good. i wonder does B chip always better than A? cuz i paid $40 more on the B chip, i tend to lead on keeping the A chip is results are not significant due to the price difference i paid.

 

[coldpower27]

How much n7 did you pay for your E6600's?

 

[The Sauce]

Well My F stepping is a true dog. It needed 1.5V to get to 3.2.

 

[n7]

Originally posted by: CloE
nice B chip, I also brought 2 e6600. only can keep one, I have L630B and L632A. they are in UPS track now, dont know which one it good. i wonder does B chip always better than A? cuz i paid $40 more on the B chip, i tend to lead on keeping the A chip is results are not significant due to the price difference i paid.

Actually, both could be very good.

If you look thru the C2D OCing databases @ XS & HF, you'll see both A & B steppings doing very well.
http://www.xtremesystems.org/forums/showthread.php?t=127379
http://www.hardforum.com/showthread.php?t=1138241

It does seem that B chips have a lower VID though, at least from what i've noticed people mentioning.

 

[n7]

Originally posted by: coldpower27
How much n7 did you pay for your E6600's?

Too much

Got the first one from DirectCanada. $373.36 + GST.
Second one i went into a local shop, A-Computers, & looked at the two boxes they had.
They had one A & B, & i grabbed the B. $378.90 + GST.

 

[n7]

Originally posted by: Snatchface
Well My F stepping is a true dog. It needed 1.5V to get to 3.2.

Ouch...that's not good at all.

You can see what mine did, but that's with my Freezone at max (i.e., uber loud) settings too.

I've seen only one good result from the F stepping thus far; everyone else has had subpar results :/

 

[lopri]

Did you push the thing @1.50~1.55V yet?

 

[n7]

Originally posted by: lopri
Did you push the thing @1.50~1.55V yet?

Believe me, a part of me is dying to go higher in a hurry, but i've decided to go up slowly this time.

Also, temps are a lot higher with this B one for some reason...not sure if it's because of using closer to the correct amount of paste, the chip is hotter, or my room temps are higher...or likely a combination of all three.

Will this keep you satiated for now?
http://img71.imageshack.us/img71/1461/e6600b3501ghz9x38913625fm2.jpg

3501 MHz w/ 1.3V load

All i can say is that this B stepping is like another better world compared to the F.

 

[lopri]

Yeah that'll do. It's amazing that you achieved such high clocks with such low voltage. But I still want to know how high it'll go with the aid of 1.50V.

 

[n7]

Well, i won't be back home till later tonite, but i'm finding some issues.

Orthos has yet to error on me, but i have had a number of BSODS thus far...oddly enough, the error reported is an ATi driver error...so i'm not sure what is going on.

Gonna have to figure that out before going any further...

 

[Brunnis]

L632B173 seems similar to my L631B120. I'm currently testing it at 3.2GHz, trying to find the lowest vcore it needs for this frequency...

 

[CloE]

nice chip n7, you found the B chip to fill up your OC greed. It appears to me I should also keep the B chip, lower voltage and higher clock? but extra $40 bucks for B stepping make me feel guilty

however i notice the 1.3voltage is from CPU-Z. usually that's not accurate after vdroop messing it up. should read from the bios voltage number

 

[n7]

Originally posted by: CloE
nice chip n7, you found the B chip to fill up your OC greed. It appears to me I should also keep the B chip, lower voltage and higher clock? but extra $40 bucks for B stepping make me feel guilty

however i notice the 1.3voltage is from CPU-Z. usually that's not accurate after vdroop messing it up. should read from the bios voltage number

CPU-Z reads vcore accurately up till about 1.43V.
Asus PC Probe 2 reports identical voltages to CPU-Z also.

Vcore set in bios is always higher than actual [vdrop], & going from idle to load has an additional drop [vdroop]

It appears Intel is starting to bin the C2D a bit now, as we have more crappy "F" steppings, & true Allendale E6400/6300s now too.

Likely all the good ones are going to the future E6230/6420, etc.

 

[n7]

Originally posted by: lopri
Yeah that'll do. It's amazing that you achieved such high clocks with such low voltage. But I still want to know how high it'll go with the aid of 1.50V.

You should be proud...i think i managed to f*ck up my XP install

Ran 3807 MHz (9x423) 4:5 & got 13.484s in SuperPI 1M.
http://img221.imageshack.us/img221/6517/superpi1m13484se6600380ox3.jpg
Then tried 3906 MHz (9x434) 1:1 for 13.219s.
http://img207.imageshack.us/img207/5951/superpi1m13219se6600390kr2.jpg

Of couse, i then figured i should try dropping teh multi so i could get my RAM speed up & timings down (8x488) 1:1 4-4-3-4.
Normally, my RAM can do that, but apparently, not tonight...

That's when things got buggered up.

As impressed with 3.9 GHz suicide runs as i am, i am not happy w/ temps at all.

Gonna have to remount the block i guess, since they're ridiculously high compared to the "F" stepping at the same vcore.

I won't be able to reach 4 GHz, since that'd take more vcore than i'm willing to put into this baby (1.55V under load is the max i'm willing to do), but yeah, fun adventure regardless

Now i have to figure out what 24/7 settings i want to use...likely gonna be around 3.5 GHz, since i can keep vcore so low.

 

[lopri]

I think 3.50~3.60GHz with C2D is like 2.70~2.80GHz with A64. Good chips will do with lower voltages, but it seems like past 3.50~3.60GHz, it gets difficult to keep stability with reasonable temps. Being able to run those frequencies with low voltages is a very good thing, though. I don't understand how AT manages to do 3.90GHz @50C load. (check recent HSF reviews)

 

[The Sauce]

Originally posted by: n7

Originally posted by: lopri
Yeah that'll do. It's amazing that you achieved such high clocks with such low voltage. But I still want to know how high it'll go with the aid of 1.50V.

You should be proud...i think i managed to f*ck up my XP install

Ran 3807 MHz (9x423) 4:5 & got 13.484s in SuperPI 1M.
http://img221.imageshack.us/img221/6517/superpi1m13484se6600380ox3.jpg
Then tried 3906 MHz (9x434) 1:1 for 13.219s.
http://img207.imageshack.us/img207/5951/superpi1m13219se6600390kr2.jpg

Of couse, i then figured i should try dropping teh multi so i could get my RAM speed up & timings down (8x488) 1:1 4-4-3-4.
Normally, my RAM can do that, but apparently, not tonight...

That's when things got buggered up.

As impressed with 3.9 GHz suicide runs as i am, i am not happy w/ temps at all.

Gonna have to remount the block i guess, since they're ridiculously high compared to the "F" stepping at the same vcore.

I won't be able to reach 4 GHz, since that'd take more vcore than i'm willing to put into this baby (1.55V under load is the max i'm willing to do), but yeah, fun adventure regardless

Now i have to figure out what 24/7 settings i want to use...likely gonna be around 3.5 GHz, since i can keep vcore so low.

You're thinking AMD overclocking. You shouldn't drop the multi on a core 2. It forces the northbridge to overclock and thats probably why you ran into instability. Unless you have an ES processor you should always leave the CPU multi at default.

MCH clock = (default multi)/(set multi) x fsb

SO initially you were running your NB @ 423 MHz

after decreasing your multi it was running @ 549.

That likely contributed to your instability. In that one, seemingly simple move, you overclocked your NB 25%. Don't worry as much about running your ram faster. Given that you were using a P5B and the strap changes at 401 you were loosening your MCH latencies by increasing the FSB anyway, so you were probably downgrading your memory performance by increasing the FSB...the opposite of what you were trying to do.

If you really have to overclock your memory then change the divider and test, but leave the CPU multi alone.

 

[n7]

Originally posted by: Snatchface
You're thinking AMD overclocking. You shouldn't drop the multi on a core 2. It forces the northbridge to overclock and thats probably why you ran into instability. Unless you have an ES processor you should always leave the CPU multi at default.

MCH clock = (default multi)/(set multi) x fsb

SO initially you were running your NB @ 423 MHz

after decreasing your multi it was running @ 549.

That likely contributed to your instability. In that one, seemingly simple move, you overclocked your NB 25%. Don't worry as much about running your ram faster. Given that you were using a P5B and the strap changes at 401 you were loosening your MCH latencies by increasing the FSB anyway, so you were probably downgrading your memory performance by increasing the FSB...the opposite of what you were trying to do.

If you really have to overclock your memory then change the divider and test, but leave the CPU multi alone.

I know lowering multi increases NB frequency, but i've had no trouble doing that before actually, with my E6300, E6400, or "F" E6600 even.

Maybe i'd just been lucky with the mobo, but in spite of that putting a lot more stress on the NB, i'd managed it before just nicely.

Since i won't be running 3.8+ GHz normally anyway, it doesn't really matter.

But for attempting things last night, the reason i wanted to run 8x vs. 9x is that with 9x, i was running 9x434 w/ RAM @ DDR2-868, since my RAM won't do 4:5 at that speed (DDR2-1085).
Best timings i can run at DDR2-868 are 4-4-3-4.

If i was able to run 8x488, i could then do DDR2-976 4-4-4-4, which would give me better benches overall.

But yeah, i'm gonna be running a lot slower speed for 24/7 use, likely only 3.5 GHz anyway.