knopflerbruce's K8 fun corner!

**uncle john** · 02-09-2009, 11:30 AM

Fellow be-2400 owner here.

Nice work with the cpu-z WR.

Mine is somewhat of a dud(3381 cpu-z) but I still managed to take both wprime records using air cooling.

**knopflerbruce** · 02-09-2009, 01:13 PM

Originally Posted by uncle john

Fellow be-2400 owner here.

Nice work with the cpu-z WR.

Mine is somewhat of a dud(3381 cpu-z) but I still managed to take both wprime records using air cooling.

I hate wprime, because I have only XP Pro

Vista is SO much more efficient in that bench. But once I have Server 2008, I'll try to kick you off the throne

That's a PROMISE!

I hope there are better BE's around, I'm not impressed by my chip either... But I have no idea what brisbane weeks that are good or bad, so maybe I just got a couple of

ty weeks

**uncle john** · 02-09-2009, 04:12 PM

I hope you like SuperPi1M.

I use 770/sb600, xpHomeSP3, and simple 150cfm air. I can't make it any easier for you. Go to work.

**knopflerbruce** · 05-18-2009, 01:58 PM

I'm quite busy these days with work and school, but my new SS is begging me to test some chips all the time

pi runs etc take too much time, so only suicide shots for now... (when my holiday is here, and I get my sticks from Germany there will be plenty of pi action, too

It stinks on air and water, at least... to my eyes it looks like the others (except the 1212 and the FX-62) are underperforming a bit, justt feel they should go a few MHz higher... Maybe the bad evap mount was the reason, causing uneven temps across the IHS... I'll find out later when I put them on dryice

That 1212 had a sweeeet coldbug, but sucked on air... maybe I can get a nice score on dryice, I dunno... not feeling too confident about that one.

It's also nice to break Kingpin's FX-62 WR, even if he ran 3D at his old record frequency

(no idea if I could, this unit won't take the heat at 3.6ghz+)

**VoodooProphetII** · 05-18-2009, 08:29 PM

Very impressive stuff. Well done.

**knopflerbruce** · 06-03-2009, 03:23 PM

Allright... my first REAL LN2 session!

Had to learn a few things, so it took a little while to get started... what's strange is that this setup seems to love some chips, and hate others. No idea why

But anyway... this is what I managed with 10L LN2:

3500+ Clawhammer @ 3505MHz 3rd http://valid.canardpc.com/show_oc.php?id=579350
3200+ Newcastle (939) @ 3526MHz WR (by 620MHz

) http://valid.canardpc.com/show_oc.php?id=579347
4000+ Clawhammer @ 3669MHz WR http://valid.canardpc.com/show_oc.php?id=579352
FX55 (it IS an FX55, same batch as the one stephen got years ago - even if CPUZ doesn't agree

) @ 3991MHz 4th (http://www.hwbot.org/result.do?resultId=860268 - was replaced by that "real" FX57 score...)
FX57 @ 4012MHz 14th http://valid.canardpc.com/show_oc.php?id=579354
Opty 156 @ 4030MHz WR http://valid.canardpc.com/show_oc.php?id=579345

I tried a couple of others too, but since I didn't break my personal best (despite better cooling, used dice before), I didn't make any validations.

Not too much time for superpi, pifast or wprime - was pretty hard to control the coldbug, since I had no help, especially since I had no thermal sensor, or thermos (had to pour straight into the pot

). But I did a couple of runs anyway, just to see what kind of scores I could get... so I got the FX55 1m WR (with the chip identified as FX57), and the Opteron 156 1m record, which actually is the best score of ANY non-FX socket 939 chip out there

(5th including the FX series, but a SHORT way up to 3rd place).

Opty 156 superpi1m @ 21.844s WR: http://www.hwbot.org/result.do?resultId=860269
"FX55" superpi1m @ 22.859s WR: http://www.hwbot.org/result.do?resultId=860270

I forgot to put the best score on my memory stick, as it was mislabeled on my bench drive

Had to capture it fast to avoid coldbug... if my memory doesn't play games with me, I think it was 22.515s @ 3776MHz.

**stunned_guy** · 06-04-2009, 12:49 PM

What an action, very impressive !!!!

You are the men for AMD

**Turrican** · 06-05-2009, 07:04 AM

awesome work bruce

**knopflerbruce** · 06-11-2009, 11:04 AM

Got some LN2 today as well (10L). I decided to try some AM2 chips, but my mobo hates me today, so I can barely do suicide shots... Nothing else than auto vcore gives me instant 1.8v, it hates all my 4 memory kits, and it failed to make my Opty 1224 hit 4.3GHz, despite the fact that it did 425x at 1.66v and 5-6C higher than the coldbug

Managed to get ONE validation above 4265, this one: http://valid.canardpc.com/show_oc.php?id=584303

Not too bad, highest CPUZ validated 90nm K8 result (3rd including the 65nm Brisbanes, 4th including the 6400+ WR picture) - but should've gotten #1 on this

Still, I'll try more. Also got this score on my new 6000+ (125w!):

http://valid.canardpc.com/show_oc.php?id=584305

WR by a mile, and still more in it - I THINK 4250 is possible. But the damn board went crazy before I had a chance to find out.

The last chip was a 4400+ Brisbane, and after struggling for 45 minutes I think I got the record there as well... have to check my benching drive first, when the board from HELL decides to start working again so I can copy the validations.

4050-4060 is pretty decent, was at 1.63v and still positive temps

Dunno what to test next... maybe try to grab some HWBoints for that brisbane and 6000+, I'm pretty sure I have a good chance on a few benchmarks, despite using the slow nf590 chipset (compared to 790fx), and XP pro (which sucks on wprime, at least for me).

**miahallen** · 06-11-2009, 12:38 PM

Very nice Bruce...did you get a thermos?

**knopflerbruce** · 06-11-2009, 02:03 PM

Originally Posted by miahallen

Very nice Bruce...did you get a thermos?

Nope

I can handle the dewar now, so why bother

Testing 939 now... got a couple of HWBoints, too

**BeepBeep2** · 06-11-2009, 02:28 PM

Thats just sick. Awsome stuff!

**[XC] Oj101** · 06-12-2009, 12:16 AM

OMG I hate you, now I have to change my sig

**knopflerbruce** · 06-12-2009, 02:48 AM

I think my chip is pretty good, it totally owns my Opteron 1222. If it had the same coldbug as the 1224 I would've hit close to 4.3... was like 15c between them

I wonder what your new chip, and stunned guy's will do

I guess I will test some stuff on water when i get home from work... Feels kinda useless only 12h after I benched with Ln2

But i want my HWBoints

**breakfromyou** · 06-12-2009, 06:22 AM

Wow man, congrats on so many impressive clocks!

Makes me want to get the dice pot out to see what my Phenom II has in it. :p

**tiborrr** · 06-15-2009, 11:23 AM

GJ bruce. I would be cool if you'd write coldbug temperatures next to every chip benched so people could get a feeling of how CB temperature differs from chip to chip

Best Regards,
Niko

**knopflerbruce** · 06-15-2009, 01:01 PM

Originally Posted by tiborrr

GJ bruce. I would be cool if you'd write coldbug temperatures next to every chip benched so people could get a feeling of how CB temperature differs from chip to chip

Best Regards,
Niko

I'll see what I can do

CCBCF's bug hard, my LE1620 did 320MHZ HTT at +9c but managed to nail the WR at 13c and 1.79v after one of the hardest battles I've ever had

WR LE1620: http://valid.canardpc.com/show_oc.php?id=586988

My LE1600 (also CCBCF) bugged at 13c and 365MHZ when I benched it.

CCB8F's are a tad better than the last one, I guess about 20-25Mhz more at that temperature. Didn't really try too many of those at that high HTT, though.

**knopflerbruce** · 08-14-2009, 06:48 AM

First results after the summer holiday: http://valid.canardpc.com/show_oc.php?id=654031 FX55 Clawhammer @ 4044MHz WR
http://valid.canardpc.com/show_oc.php?id=652802 3500+ Venice @ 3769MHz WR
http://valid.canardpc.com/show_oc.php?id=651044 3000+ Newcastle (s939 version) @ 3290MHz WR

Nice start, but a few broken mobos as well

**stunned_guy** · 08-14-2009, 07:12 AM

You are crazy knut

that 3000+ rocks very good.
whats the stepping of that fx-55 ? is this mine cpu ?

why broken motherboards ?

**knopflerbruce** · 08-14-2009, 08:18 AM

Originally Posted by stunned_guy

You are crazy knut

that 3000+ rocks very good.
whats the stepping of that fx-55 ? is this mine cpu ?

why broken motherboards ?

The same chips as you had, yes.

...and here is your old FX53 in action: http://valid.canardpc.com/show_oc.php?id=654224

It sucked hard on dryice temps, but once it went REALLY cold it wanted to fight it's old record

But now I've used enough Ln2 for 130nm, I guess am2 next

Broken boards because of WC failure (5L on benching rig while it was running on LN2) and broken socket (technically it still works, but it's too much hassle to remove chips so it's unbenchable.). But it still works for 24/7 (typing on it now

).

ASUS Crosshair + some Brisbanes next I guess...

**knopflerbruce** · 08-14-2009, 09:18 AM

New score... http://valid.canardpc.com/show_oc.php?id=654322 Sempron 3200+ Palermo @ 3351MHz WR

...but this one was hard, the temp sensor was all over the place. While loading in windows: 30c, idle: -25c (!).

**stunned_guy** · 08-14-2009, 12:18 PM

I wonder why our fx-55 and fx-53 runs better on your board than our board ? all two records are stolen of you

we ( gianni and me ) thought that will be records for a the future...

**knopflerbruce** · 08-14-2009, 02:25 PM

Originally Posted by stunned_guy

I wonder why our fx-55 and fx-53 runs better on your board than our board ? all two records are stolen of you

we ( gianni and me ) thought that will be records for a the future...

I can't explain it... are you sure you got the optimal temps? I know I was close, but because I had no temp sensor I doubt I was at the limit. Also, I seemed to gain 10-15mhz by going from 1.77v to 1.84v on the FX55, so maybe you used too little voltage for that one

The FX53 was really moody when it got below 145 in ITE, but on a couple of occasions it let me get past 3900.

Could be the board of course. Maybe I'll test the FX53 I got to 376x on on my Expert. I thought that one would kill your record, but it didnt

It managed 3560 or something on dryice (on a Venus, not mine), and worked all the way down to 130 in ITE, so 85 degrees lower temps should've given me more than 200MHz extra

Allright... I call it a day now

Tried a couple of chips on the Crosshair on air, maybe I'll crack the Opteron 1210 WR tomorrow

Didn't look to have a too bad CB for these CCB8F chips.

Forgot to mention that the 3000+ NC was an xPBW chip, so they're not THAT bad.

Edit: we should've done these OC's while these chips were hot... then we would've been heros

**Turrican** · 08-15-2009, 11:09 AM

awesome score bro

Thread: knopflerbruce's K8 fun corner!

Thread Tools

Search Thread

Rate This Thread

Display

Bookmarks

Bookmarks

Posting Permissions