Conroe @ Joe Camel's house

[Joe Camel]

will put pics etc up here


post away i'll try to keep y'all up to date here


i got it 8M stable 400x10 ... going for 32M

[Qubix]

What is your system specs, know you got a X6800 but what is everything else?

[traldan]

Sweet man can't wait to see it.

[hainer36]

might be a dumb question here....but is it under phase :D

[Sir. BOBSONATOR]

Step 5?

[d94]

well, that surpases the ****ty ass results @ XS
there's hope for retail conroe yet! :D
ill be watching this thread
very, very, very, very, very, very(6x) closely

[Qubix]

He mentioned in another thread using his Storm waterblock so I would assume its under water.

[Joe Camel]

ASUS P5W DH no volt mods but better cooling

Storm G5 (silver), Laing D4 pump @ 14V, twin single pass "dual-heater cores" in series, 4x 130CFM fans

Mr zippy 700w PSU







http://img477.imageshack.us/img477/1254/pict0260hy8.jpg




will edit more stats soonish


this rig but not this Vid card:



http://img142.imageshack.us/img142/7231/pict0259tb8.jpg



.

[Ross]

Step 5?Retails are Step 6 (B2).

JC, I am not sure what all you've been trying, but try these if you haven't already:
Performance Mode = Standard
PCIe = AUTO (default)
PCI Freq = 33.3MHz

FSB Termination = AUTO (little to no reason to up this and it usually hurts more than helps)
MCH Voltage = 1.65V
HyperPath3 = Disabled

Disable C1E, Thermal Control, EIST and all that stuff under CPU Config. Just leave the microcode update enabled.

[fhpchris]

Single 32m or dual?

[Gautam]

8M stable, awesome!!! Last we talked you couldn't even do it 1M. What setting(s) did you change to get such a big leap in stability?

Way to pwn the XS retail results. :D

[greenmaji]

GO JOE!!! :attn:

[xgman]

cpu temps?

[IYIENACE]

C2D E6600 on pre-order $350
Asus P5WDH on the way $250
OCZ 800MHz DDR2 memory sitting on my table $200

Waiting for July 27th... agonizing

Love your sig man, feel your pain! :santa:




(and JC subscribed~ GL)

[party animal]

Nice job JC! 4Ghz stable. Wow! Congrats! Now lets see the Dual prime! :D

[greenmaji]

4Ghz 8M superpi stable.. give him a while there may be more :D

[Jame Gumb]

Doubled up on those bonneville's with the CB shroud...4x130CFM that room must be humming! 4GHz!!! "maybe more" :drool:

[fhpchris]

8M stable, awesome!!! Last we talked you couldn't even do it 1M. What setting(s) did you change to get such a big leap in stability?

Way to pwn the XS retail results. :D

All of the XS retail results are Dual Prime stable, not 1m or 8m.

Some X6800 owners saw 55C load on high end water cooling, 1m and 8m will not stress cooling enough to show stability, and clearly only one thread will not either.

Show highest Dual Prime stable... It wont be 4ghz.

[Ross]

All of the XS retail results are Dual Prime stable, not 1m or 8m.Yes, XS is all about stable OCs, LOL.

Who started all this about "stable" anyway? Unless he went from water to phase, anyone thinking it went from being barely stable enough for 1M SPi to stable for dual 32Ms or P95s ( :rolleyes: ) is living on another planet.

[fhpchris]

I am only pointing out that Each retail X6800 @ XS did "worse" because the users wanted 24-7 rigs that were "stable"

Many people have stepping 5 CPUs that do ~4.1 on air.(Pi stable)

Andre Yang's retail X6800 did ~3.7 dual prime stable, but 4ghz + 1m/8m.

[Joe Camel]

with Gautam's setting (some a little diff from Ross's) ....

i can dance around 4.0 all i want with dividers or higher or lower multi etc etc but i cant do much better than 4.0

as for stability, im only using Pi 1m right now just to find the benching limit ill try to do some dual Pi some time soon maybe some dual prime but dont count on it being B4 the weekend.

[Jame Gumb]

Did you have to lap the stepping off?

[RangerXLT8]

Woot looking good JC!

[Stilletto]

Retails are Step 6 (B2).

JC, I am not sure what all you've been trying, but try these if you haven't already:
Performance Mode = Standard
PCIe = AUTO (default)
PCI Freq = 33.3MHz

FSB Termination = AUTO (little to no reason to up this and it usually hurts more than helps)
MCH Voltage = 1.65V
HyperPath3 = Disabled

Disable C1E, Thermal Control, EIST and all that stuff under CPU Config. Just leave the microcode update enabled.

A little off topic here, but wouldn't disabling thermal control be a little dangerous? Wouldn't your processor not throttle to protect itself?:confused:

[Ross]

Oops, that's for phase/DI. It shouldn't matter anyway unless you plan on running your proc at 80C without paying attention ;)

[greenmaji]

Andre Yang's retail X6800 did ~3.7 dual prime stable, but 4ghz + 1m/8m.

Andre Yang has a huge pile of E6600 stepping 5 ES's for sale.. Now why do you think his retail processor did so poorly HHHhhhMMMMMmmmm
And since he has a huge pile of ES's why do you think he bought and benched (poorly) a retail processor when he could have got his inside guy to get him an X6800 for REAL cheap?? HHHhhhhMMMmmm

I would like to forget about that thread ;)

BTW.. were did you see anyone 4Ghz dual P95 stable on air over at XS, please post the link ;)

[fhpchris]

Andre Yang has a huge pile of E6600 stepping 5 ES's for sale.. Now why do you think his retail processor did so poorly HHHhhhMMMMMmmmm

He took a loss on his X6800 also, he sold it for less than 1100$ US.


And since he has a huge pile of ES's why do you think he bought and benched (poorly) a retail processor when he could have got his inside guy to get him an X6800 for REAL cheap?? HHHhhhhMMMmmm

Andre's results match everyone elses results just perfectly, he was not sandbaging.

Andre clearly did 4ghz 1m.

Besides, some people think that the retail steppings might end up rather good.
I would like to forget about that thread ;)

BTW.. were did you see anyone 4Ghz dual P95 stable on air over at XS, please post the link ;)

4.1ghz on air isnt prime stable, the highest prime stable air/water clocks I have seen are from Andre's E6600 Step 5 and Mdzcpa.

it bombed at 1.5v about an hour in, but than ran all night at 1.525v @ 3.9Ghz.

That is a x6800 Step 5. I thought I saw a prime @ 4ghz from another user, but it was not over a long period of time(8 minutes?), and it was also on high end water...

Andre's says his E6600 Step 5 does 3.9ghz dual prime also.

Most E6600s-x6800s are doing 3700-3850 prime stable.

[Joe Camel]

.


single 32M stable @ 4.0GHz 400x10 1:1 3-4-3-4



http://img236.imageshack.us/img236/5157/32m13m8s40zn5.jpg




trying DUAL 8M (looks good.... NOPE :()

will host pic (not if it doesn't finish :()




Gautam, make sure you try FSB termination volts @ AUTO


.

[Gautam]

Looks like you got one weak core. :-/

What did you change to get it from barely able to run at 3.95 to 32M at 4.0??

Btw, update your CPU-Z versie.

[Gautam]

Sit and think for a second about just how much of a killing he's making off of each ES he sells. (Hint, over 100% return) Losing even a couple hundred bucks on a single sale is just a teardrop in a lake.

He took a loss on his X6800 also, he sold it for less than 1100$ US.



One of my friends with a step 5 6700 (a VERY good one) can do 1M at 4 GHz but last I heard having trouble getting even 3.7 to run dual prime for more than a minute. These retai XE's just might be worth their salt. My interest, personally, is rekindled.

[Joe Camel]

too many things to remember sorry :(
(hyperpath3 = OFF and 1.66+ Vcore are probably the biggest though)

im basically @ your suggested settings except with Ross's FSB term volts = AUTO

this thing is NO "White Box" retail 6800 thats for sure but... i guess we'll see how it likes the cold and then pass final judgment


tightened the mem "some" and running 32M again.... 400MHz @ 3-3-2-1 :D

[d94]

:(

[Joe Camel]

http://img211.imageshack.us/img211/1787/tight32mqx5.jpg

:D:clap::clap::D

[Gautam]

JC, don't be disheartened by coolaler's results. We've talked about this before, those guys are literally impossible to keep up with. He handpicks CPUs literally out of HUNDREDS. Every result you see from him is heavy pre-screened. It's not like he clicks onto newegg and buys a single proc and gets lucky...I hear he's a major distributor himself, so he has access to as many samples as he wants. Every one that he shows publicy is a golden one. Pedro Rocha's is probably a similar case. Don't forget he was the first one to ever break 4 g's with an AMD...it takes a lot more than dumb luck to do that. I'm crossing my fingers for 4.4 on your Chilly1 SS. If it can do that, it'll make it just about the fastest on single stage for everything except Pedro Rocha's and coolaler's white boxes. Like I said, yours looks about 200MHz or so faster than my E6700 which is one of the notorious step 5's. I would say its a step forward for sure and most likely I'll still be purchasing one of these.

[Gautam]

http://img211.imageshack.us/img211/1787/tight32mqx5.jpg

:D:clap::clap::D
Awesome, you just tore my 32M time a new one. :D

You and ROSS need to start posting your results in the Pi threads NOW.

[jpersinger]

Joe,

whats the mem? Also, can you push the FSB any further than 400 MHz? Thanks.

[Joe Camel]

Corsair TWIN2X2048-6400C3 (3-4-3-9 stock)

max ive been able to push the FSB is 442 but wasnt quite (as) stable
(no V-mods on mobo)

[xgman]

Joe, again, what temps are you seeing and from what source please?

[Joe Camel]

haven't even looked/tried

im letting stability guide me

if i let AI booster load, it crashes the rig EVERY time

ill get a few different programs to see where im @ after work tonight



IIRC right as AI-booster starts to load, (B4 i removed it) ive seen 38C

[crimedog]

freeze it! volt mods! freeze it!

[Joe Camel]

i hear ya :p

have you looked @ these damn mobos and how close they are putting the caps to the socket?? its going to be lots-o-"fun" to insulate :rolleyes:

cant decide on V-mods yet: V-DMM and MCH?

ill probably have "had enough" on water by this weekend and will start work on the insulation

[crimedog]

grease, neoprene + exacto knife :)

i'm sure guatam knows more about the vmods for these boards than i do but i'd definitely want vdroop and vcore

[xgman]

haven't even looked/tried

im letting stability guide me

if i let AI booster load, it crashes the rig EVERY time

ill get a few different programs to see where im @ after work tonight



IIRC right as AI-booster starts to load, (B4 i removed it) ive seen 38C


Might try http://www.xtremesystems.org/forums/showthread.php?t=103638&highlight=coretemp

or Asus pcprobe 2

What does your hardware monitor say in the bios for cpu temp?

[RangerXLT8]

grease, neoprene + exacto knife :)

i'm sure guatam knows more about the vmods for these boards than i do but i'd definitely want vdroop and vcore

Proth said he was going to do the droop mod(i imagine he is busy with RL) before the weekend is done.

[xgman]

Joe, how tight are your thumb screws tightened down on your G5? On my first P5WDH I think I overtightend and messed up some of the pins in the socket. Also which way did you mount the G5? towrads the back of the board lengthwise or towards the side?

[Joe Camel]

side to side if our terms are = ...

the in/out barbs are next to each other with one closer to the mem and one closer to the IO/back of mobo

i crank them sucka's down TIGHT on a CPU with IHS but try not to bow the mobo too much ;) ;)

thanks for the temp probe suggestions, i was thinking speed-fan and the one that read Merom/Yonah temps...(cant rem name)

isnt there a Intel one which shows a bunch of volts and core loads and mem use??? eva2000 on XS used it ??
http://www.xtremesystems.org/forums/showpost.php?p=1589045&postcount=18

or would that be intel mobo ONLY?

[Ross]

That app only works on the BadAxe. There's not too many things that I trust for temps on Conroes yet...including CoreTemp.

[xgman]

side to side if our terms are = ...

the in/out barbs are next to each other with one closer to the mem and one closer to the IO/back of mobo

i crank them sucka's down TIGHT on a CPU with IHS but try not to bow the mobo too much ;) ;)

thanks for the temp probe suggestions, i was thinking speed-fan and the one that read Merom/Yonah temps...(cant rem name)

isnt there a Intel one which shows a bunch of volts and core loads and mem use??? eva2000 on XS used it ??
http://www.xtremesystems.org/forums/showpost.php?p=1589045&postcount=18

or would that be intel mobo ONLY?

That's intel MB only. I tried anyway, lol. There is also a good MEmeset progy for this on the X Intel page stickies.

So you mounted the short side of the G5 along the side of the socket mounting bracket arm? Or the short side along the Mofsets copper heatsink side. When I tried the mount towards the heatsink it was extremely tight under there and touched the sink and maybe on top of the caps under there.

[Gautam]

side to side if our terms are = ...

the in/out barbs are next to each other with one closer to the mem and one closer to the IO/back of mobo

i crank them sucka's down TIGHT on a CPU with IHS but try not to bow the mobo too much ;) ;)

thanks for the temp probe suggestions, i was thinking speed-fan and the one that read Merom/Yonah temps...(cant rem name)

isnt there a Intel one which shows a bunch of volts and core loads and mem use??? eva2000 on XS used it ??
http://www.xtremesystems.org/forums/showpost.php?p=1589045&postcount=18

or would that be intel mobo ONLY?


I'd say pull the Fluke out, that's the only reliable method. :-/ The Bad Axe had a particularly weak PCB. I dunno if the Asus is the same way, but I use the mounting bracket from my Tuniq Tower as a backplate for both my Prommy and Chilly1 container. Gives plenty of peace of mind as I crank things down.

I have yet been too lazy to even take DMM readings, but according to Ross vdroop isn't necessary, isn't that right?

VDimm I think is virtually useless and even possibly dangerous for anything but D9 fatties. D9GMH and whatever else need not apply. I was talking with a friend about this early. The fatties can take a heavier beating than anything else it seems. He happened to have an Infineon stick (I think) and a Mushkin D9 fatbody stick in a dead Bad Axe too. One stick survived, the other is dead. Guess which ones which. The Bad Axe gave my 5400UL over 4V when it "popped" and they're still living whereas Ross' old 6400C3 met its end, even though both of our little disasters seemed identical. I don't know if even *I* would personally toy around with much more than 2.4v with the 6400C3's, or for that matter anything other than D9DQx.

Vmch might be helpful, but less so with an XE. From what I've observed, the best results seem to come from running the FSB in the high 300 to low 400 range and then running 5:4. Ross pulled out some stellar Pi results in this fashion if I'm not mistaken, and its how some of the quickest Team Japan scores were achieved as well. So in short, more likely than not the FSB will be fast enough unmodded. 11x multi should do the trick probably.

And I guess that does it for vmods...chances are you won't need any. (Sorry Dan :p)

Vcore in all its oddness on the board, might be nice to have a pot to control that though. The fine tuning if nothing else, might be helpful.

[Joe Camel]

X-Gman
ill take a pic when i get home (15 min or so) and post it here ;)



Gautam,
so then i get the temp reading of the edge of the IHS... ill fly blind B4 i spend time to know THAT temp :p

[Gautam]

Edge of the IHS would still be better than the other readings IMHO. Should be pretty close to actual die temp, a little lower of course but would give you a decent idea of how much vcore is affecting temps, etc.

[Ross]

That works pretty well and is definitely better than nothing or something unreliable.

http://i10.photobucket.com/albums/a125/boostedevo/dice/probe.jpg

[Killaapp]

Ross - how accurate do you beliebethat is? After seeing that picture, I want to rig something similar to that.

[Joe Camel]

sorry bats in digi-cam are dead....charging

here is how its mounted though:

(this is an Aopen mobo with a Merom NOT the ASUS with Conroe)
http://img102.imageshack.us/img102/7315/pict0012vc8.jpg

[Ross]

Ross - how accurate do you beliebethat is? After seeing that picture, I want to rig something similar to that.Pretty accurate I think...I used that for DI and it was pretty close what we were expecting. It doesn't move as fast as an on-die temp monitor obviously, but it will definitely let you know when it's getting warm ;)

[Joe Camel]

IDLE


http://img56.imageshack.us/img56/9006/idletempbi0.jpg



"1/2" load

http://img295.imageshack.us/img295/120/halfloadtempsfj3.jpg



"full" load


http://img295.imageshack.us/img295/4051/fullloadtempiq5.jpg

[party animal]

wow, these Conroes get pretty toasty dont they?

[RangerXLT8]

1.63v, keep it up and that 1000.00 chip won't last too long. [Edit] At any rate, they are some impressive numbers!

[party animal]

On the topic of votages, if we were to run Conroes on water, what would max vCore be?

[dylskee]

@RangerXLT8, That 53c temp is 100% load, his idle temps are 35 from what i can see. ^^^ He is on water party animal.

[greenmaji]

I wouldn't call it 100% load but he's got some load on it.

[dylskee]

I wouldn't call it 100% load but he's got some load on it.
Post #56, third screenie shows 100% load, no???

[Gautam]

SuperPI isn't quite a 100% load, regardless of what monitors might show. Other apps such as Prime and OCCT drive temps higher and stress the system out more. SuperPI actually doesn't even deliver a constant load, it goes up and down on each iteration.

[party animal]

@RangerXLT8, That 53c temp is 100% load, his idle temps are 35 from what i can see. ^^^ He is on water party animal.

Yah, but for 65nm processors, what would be ideal voltage on water for 24/7 vCore?

[dylskee]

Yah, but for 65nm processors, what would be ideal voltage on water for 24/7 vCore?
Well as far as the Conroe i have no idea, but my Cedar Mill i wouldn't run more than 1.5v. Some people run as high as 1.55v, i already killed a 661 Cedar Mill so i'm a little gun shy now so i run 1.3125v on my setup.

[party animal]

cool, thanks for the info!

[RangerXLT8]

wow, these Conroes get pretty toasty dont they?

Conroe doesn't get toasty my friend. Intel Pentium D 9xx[Presler] get toasty. lol Seriously I had to back down the OC because it was heating up my room.

I can't wait untill my Allendale comes to my door. This 920 will be yanked in about 2.2s.

[party animal]

Well, seeing that JC loads @ 51*C on water and not even getting 4Ghz, it seems like these Conroes will have a hard time getting to 4Ghz on water 24/7.

[ochungry]

Well, seeing that JC loads @ 51*C on water and not even getting 4Ghz, it seems like these Conroes will have a hard time getting to 4Ghz on water 24/7.
Anad "anadtech.com" says he did 4ghz on air.
I like to see the retail's do what anad says. If not, anadtech.com will be flooded w/ complaints. His credential damaged, and the trust lost if it turns out that his Engineering Sample was one of a kind.

[Gautam]

The Press Kit's that Anand and others receive appear to be markedly better than retail. :(

This is Intel's doing only, Anand couldn't have known better. In fact everything is very blurry right now, this is just my own pessimistic opinion.

[party animal]

If its true, there will a lot of unhappy oc'ers.

[orionlion82]

Well, seeing that JC loads @ 51*C on water and not even getting 4Ghz, it seems like these Conroes will have a hard time getting to 4Ghz on water 24/7.

i beg to differ. joe camel knows his stuff. he knows that much lower temps on water is possible, but his hardware does not match this. most watercooling setups dont get -significantly- (emphasis) better than high end air.
there are a few that do, but they are either high end OR arent your everyday water rigs.
thing is most people dont want to deal with exotic watercooling.
i can understand how it works with joecamel - and most others too. they dont want to do plumbing - they want to overclock, and we want to see the results.
and theyre just gonna go on phase/di at some point anyway, so why put that much into it?
it is my belief now that the most exotic/high end watercoolers MIGHT get 4ghz.
its all about heat dissapation.
i run 1.6V too. ive got a prescott here - but still my load temps are sub 40C
i designed my system to move FAR MORE heat than a computer could ever concievably generate. exotic? yes! would i bet on my odds of geting 4GHZ on conroe if i had it? yes! would the average watercooler ever go through the misery of inventing their own cooling system? never. will there be a lucky few who do get 4 on average water? i sure hope so, but sadly, low end and average water systems - may not see that magic number. dont get me wrong, i want more gigahertz for everybody, but only a few watercoolers will see 24/7 stable OC's at or above 4 on water - i just wanted to clarify your point really. its not as impossible as it sounds, allthough it wont be as common as all the birds in the sky.

[party animal]

IMO, JC's setup right now is pretty good, two heatercores in series, 4x130CFM deltas and a silver G5. That sounds pretty high end to me. But it all depends on the ambient temps in the room that JC has this set up in and it depends what ambient temps that the average watercooler's ambient temp is too. But i agree with you that it won't be a common thing.

[pscout]

Joe, Glad to see it arrived when it was supposed to :)

why not run a couple of instances of that old qmd for 5-10% just for the heat of it :D

that will tell the tale on stability and max temps.

[orionlion82]

Anad "anadtech.com" says he did 4ghz on air.
I like to see the retail's do what anad says. If not, anadtech.com will be flooded w/ complaints. His credential damaged, and the trust lost if it turns out that his Engineering Sample was one of a kind.

i would blame intel, not the reviewer.

in which case, whatever chip, whatever clocks, all that is all worth nothing if its a "special" chip when its reviewed.
this allways happens though, no?

not this time, not this time, not this time, no not like last time, or the time before that or...
we allready know the boards _SO FAR_ are kinda iffy... par for the course this early on, but still a touch dissapointing.
now if the chips dont deliver - (and the speculation ive seen around was that the real retail chips would be BETTER than the es's - a "not this time" for sure)
this is why this community is so valuable. the pioneers set, make and break expectations that everyone else operates under.
so far its gone from "too much hype">"not really any hype - WOW!">"some hype"

and plus, i trust JC more than anand :santa:

[pscout]

given intel's recent earnings results, I would not be a bit suprised if intel made sure the early reviewers got hype'able results.

But I expect reality for us mainstream OCers should not be too far off some of the early results. I just hope i get an early 6600 to find out for myself and am not disappointed. And that the retails all OC consistently.

[Joe Camel]

this is the BEST water can do (IMHO)
running @ 1.6625V (BIOS) will heat that lil' chip up no matter what you cool it with... well almost anything... ;)


insulation is 99% done:



http://img57.imageshack.us/img57/4232/pict0261bj9.jpg




tomorrow (after work) ill put the Chilly1 Single Stage on it and see how she rolls :burn:



.

[RangerXLT8]

BE looking forward to your results JC!

@ everyone else: JC WC setup is about the best you get, without chilled water\very very cool ambient. I seriously think that to even expect 3.7ghz stable is a farcry for WCing. I think that higher-end E66-6700 will top out around 3.4ghz STABLE 24/7 on water and the lower end E6400 at about 2.93ghz. This is all in terms of 24/7 stable.

[-FX-]

pretty bum :P 4ghz should have been a given :(

[party animal]

Ranger, what would say would be the max load temps for Conroe running 24/7? Should I keep vCore under 1.5V for 24/7 usage.

@JC: I'm looking forward in seeing your results. This is going to be interesting :D

[Ross]

Good luck on the SS. Here's hoping for >4.4GHz :beer:

[proth]

i hear ya :p

have you looked @ these damn mobos and how close they are putting the caps to the socket?? its going to be lots-o-"fun" to insulate :rolleyes:

cant decide on V-mods yet: V-DMM and MCH?

ill probably have "had enough" on water by this weekend and will start work on the insulation

Rich, I actually moved the caps to the backside of the board, Now there's a good 1/2 in wide area around the socket for insulation. Also cut off the "hook" on the bail that holds down the lid for easier insulation.

Good luck man !!

[greenmaji]

Just for the water peps.. what was the ambiant temps in the dungen Joe?

And good luck with the Chilly1! :attn:

Proth moved the caps to the other side of the mobo :rolleyes: I'm not supprised ;) Carefull Proth or you might end up getting legend status with those hardmoding skilzz :p

[LandShark]

subscribed :cool:

as Pete said, try to run some ancent QMDs to see some "real" load temp. honestly, I'm a bit disappointed with your temp (highend WC there!). that could mean highend air at high 3Ghz will be again near 60c under 100% load.... (I sure hope I'm wrong!!) .....

[greenmaji]

If all his F@H rigs were running the ambiant might have threw his high end water for a loop ;)

[Joe Camel]

room temps were near 75F



proth,
now THATS dedication :beer:



i put a layer of dielectric down first than a layer of 1/4" foam cut to fit around/between all the caps etc (under the lock arm)
then filled in the seams/gaps with more grease then put a layer of foam tape over that (pealing the one piece back in the pic so you could see the lower layer)
that brought the insulation layer up to and a little beyond the plane of the IHS. now it up to the Chilly1 insulation to take it from here.


here is that pic @ full size (if you wanted a closer look)
http://img74.imageshack.us/img74/7901/pict0261rl5.jpg

[Gautam]

Looking good, I'll mimick that insulation. I just had vaseline squirted all over everything...this is cleaner and probably a bit safer too. :p

GL with the XE!!

[fhpchris]

Conroe doesn't get toasty my friend. Intel Pentium D 9xx[Presler] get toasty. lol Seriously I had to back down the OC because it was heating up my room.

I can't wait untill my Allendale comes to my door. This 920 will be yanked in about 2.2s.

Some people on XS say that @ high voltage on air, the prescotts they own run COOLER than Conroe.

Conroe uses alot of power, and creates alot of heat when the voltage goes up.

;)

[xgman]

sorry bats in digi-cam are dead....charging

here is how its mounted though:


Thanks, I have mine mounted the other way. The way you have it on the asus was right up against the mofsets heatsink, but it seems to be working fine for you.

[xgman]

[IMG]http://img56.imageshack.us/img56/9006/idletempbi0.jpg


Now your temps seem more like what I'm seeing. What are your bios settings for this run as far as multi, fsb and mem speed to get the 5xx speed?

[Gautam]

Yeah even if you don't remember the exact changes, list all your important BIOS settings here please...it'll help all of us (and you if/when you reset your BIOS :p)

[xgman]

As far as voltage, I wouldn't put more than 1.6 and only if minimum of good water. maybe 1.5v 24/7 on water. As far as retail overclocking on water. The 6800 retails I've seen including my own, will do close to 4 gig but 24/7 stable , more like 3800 ish will be the norm I think. I think the majority of the ES smaoples were juiced in my opinion, even though some of them were not as good as others. I have yet to see a super duper amazing retail conroe, not that there is anything wrong with a conroe at 3700-3800. ;) At this pint I think we need better boards for the Conroe.

[Joe Camel]

Now your temps seem more like what I'm seeing. What are your bios settings for this run as far as multi, fsb and mem speed to get the 5xx speed?
well, that IS @ idle

anyway... 5xx speed? (mem speed?)

assuming thats it:
uggg i could try to put some #'s down but since im @ work they might not be right and i dont want to mislead you.



GUESSING:

11x (CPU multi)
whatever one notch over 1:1 is (mem divider)
2.4V (mem)
5-5-5-10* in BIOS (mem timings)


*could lower to 4-4-4-4 IIRC but could NOT boot @ that

1.6625 Vcore
max MCH
AUTO FSB termination
1.2v ICH
hyperpath3 OFF
(can keep the other chipset setting @ AUTO, Gautam)

will add more need to do some "work" now :p



back

ummmmm think that about covers it... if you can think of more list them and ill try to remember :p


side note:
even with the WAY higher mem MHz, the looser timings put my (sandra) bandwidth lower than when i was @ (1:1) 400+, 3-3-2-1 so im back on that

[Gautam]

Don't use Sandra to compare, it sucks. Use SuperPI 8M to figure which one's really faster. But it wouldn't surprise me. That CAS 3 sure is a killer. I just might have to buckle down and buy a set of these, it seems resistance is futile. :(

[Joe Camel]

my 32M run was on them TIGHT timings and i beat your time when you were @ higher CPU MHz so.... resistance IS futile :p

[Gautam]

Well that was from me having only 2x512...was thinking of getting a cheaper 2 gig set and running 5:4 for 500 4-4-4-4, thought maybe that might be more competitive.

[Joe Camel]

forgot about 2x gig boost :-/

arnt there some "team" 2x gig sets with cas3's?
@ like 1/2 the price of these Corsair's


EDIT:


ooooOOOOO 3-3-3-8
http://www.tankguys.biz/ddr2-3338-p-1676.html

[Gautam]

CAS 4 actually. Half the price of the Corsair yeah, basically. 500's 4-4-4-4 even 4-4-3-4 they'll do...400's CAS 3 probably impossible. I've got limits (high ones I'll admit), you don't. :p The $260 that I spent on my 5400UL's was already the most I ever spent on ram...$450-500 just makes me shudder. I'd actually be willing to take a hit of several seconds in 32M to save a couple hundred bucks. :p

Edit- check THIS (http://www.xtremesystems.org/forums/showpost.php?p=1593024&postcount=367) out. :drool: That set costs only $190!! Might have to buy two. ;)

Triple edit- I was talking about their 6400 sets. Those only come in CAS4. The $190 set I'm talking about is 3-4-4-8...cheap enough to make me wanna take the gamble...

[Joe Camel]

hehe ill try and see if this rig can take all 4x gig's of this stuff (Corsair) after i see how the SS plays out on 2x gigs


man i think im coming down with something :( ... cough cough... might have to take a 1/2 day sick/vacation ... cough....cough :o

[Gautam]

If you could, compare 500 4-4-4-x against 400 3-3-3-x in SuperPI 8M or higher sometime. Should have plenty of time now that you're "sick". :p

(You are joking, right?) If not, hope you get better soon.

[toddm27]

CAS 4 actually. Half the price of the Corsair yeah, basically. 500's 4-4-4-4 even 4-4-3-4 they'll do...400's CAS 3 probably impossible. I've got limits (high ones I'll admit), you don't. :p The $260 that I spent on my 5400UL's was already the most I ever spent on ram...$450-500 just makes me shudder. I'd actually be willing to take a hit of several seconds in 32M to save a couple hundred bucks. :p

Edit- check THIS (http://www.xtremesystems.org/forums/showpost.php?p=1593024&postcount=367) out. :drool: That set costs only $190!! Might have to buy two. ;)

Triple edit- I was talking about their 6400 sets. Those only come in CAS4. The $190 set I'm talking about is 3-4-4-8...cheap enough to make me wanna take the gamble...
just to let you know I bought that set of teamgroup 667 and mine do up to 530 4-4-3-16, with 2.2v memtest stable, great value ram:bday:

[Joe Camel]

If you could, compare 500 4-4-4-x against 400 3-3-3-x in SuperPI 8M or higher sometime. Should have plenty of time now that you're "sick". :p

(You are joking, right?) If not, hope you get better soon.
can DO :thup:


for the most part, i never really get sick (maybe 1x time per year) so to make...cough...up for that, i ...cough... "get sick" every now and ...cough... again :p

thanks for the concern though :D

[Gautam]

just to let you know I bought that set of teamgroup 667 and mine do up to 530 4-4-3-16, with 2.2v memtest stable, great value ram:bday:


Ok, that did it for me. :D I'm not quite at JC's level of amassing memory just yet but I'm learnin'. ;)

[toddm27]

Ok, that did it for me. :D I'm not quite at JC's level of amassing memory just yet but I'm learnin'. ;)
:eek: cool, is 4gb's :drool: really needed tho, haha what was I thinking I may have to order another set,

jc - nice results on the 6800, making me wish I had mine in my hand, 6700 in a week hopefully:santa:

[mkorpal]

Ok, that did it for me. :D I'm not quite at JC's level of amassing memory just yet but I'm learnin'. ;)

I just ordered mine. Looks like they will all be gone in a few days.

[TankGuys]

Those 3-4-4-8 Team group sticks are all sold out now :(

I don't think Team is making any more, either. They are being replaced by the 3-3-3-8 sticks instead. They are more expensive, $247 or so, but that's still a nice price for such good memory.

Sorry we couldn't keep the other stuff around longer. I think we sold a few hundred sets... it moved quick!

[xgman]

side note:
even with the WAY higher mem MHz, the looser timings put my (sandra) bandwidth lower than when i was @ (1:1) 400+, 3-3-2-1 so im back on that


i'm with you on this, but maybe max 3-4-3 or even 4-4-4 , certainly not 5-5-5.

[party animal]

Finally, Tankguys have the 3-3-3-8 Team Group memory listed. Now time to play the waiting game. Can't wait for the numbers to pop up JC! :D

[Gautam]

Those 3-4-4-8 Team group sticks are all sold out now :(

I don't think Team is making any more, either. They are being replaced by the 3-3-3-8 sticks instead. They are more expensive, $247 or so, but that's still a nice price for such good memory.

Sorry we couldn't keep the other stuff around longer. I think we sold a few hundred sets... it moved quick!
lollers..hope you don't mind me posting this but its not often I get lucky with anything. :D


The comments for your order are

Gauntam-

You're lucky - you got the last two :)

These should ship tonight, and we'll get you the tracking info once we have it. If you need anything in the meantime, let us know!


Good thing I didn't sit around debating for another night like I was planning. (If you recall I've been *interested* for the past month or so but never had the time to pull the trigger) I've always been "behind" in terms of amount, when most ran 512, I ran 256, when everyone was running 1 gig, I was running 512...now I'm running a gig...well that's changing now. :p

Anyways I appreciate it TankGuys. It's good to have you around to provide a retailer's perspective in the community.

[mkorpal]

Well, that's a bummer. I guess I'll have to figure something else out. I really don't think the new 3-3-3-8 memory is going to be worth $50 more. Especially since I really don't have much money to spend in the first place. Oh, and Gautam, you got so lucky.

[TankGuys]

Yea, he did get lucky :D

The new stuff is more expensive, obviously - but like I said, it's still a nice price. As far as I know, there isn't any other memory of that performance near that price point. Hopefully though, memory prices in general will start dropping once the Conroes get more widely available.

[party animal]

didnt OPB say that the chips used in the 3-3-3-8 set were the same ones used in Corsair's high end RAM? Just binned lower?

[Gautam]

They all use D9GMH chips AFAIK. The Corsairs are just binned EXTREMELY aggressively.

I'm still skeptical as to whether these sticks can really pull off speeds like that with timings like that...that post on XS just shocked me, but I'll find out first hand soon enough. Performance like that already puts it well within 50MHz of the Corsair at both the 3-3-3-x and 4-4-4-x level...and for well under half the price. Might just give the 3-3-3-9 5400 stuff a whirl too down the road. :p I'd imagine they're tested a little more vigorously.

[party animal]

Doesnt hurt to try eh Gautam? :p I might pick up a set of these myself when i buy my E6600. But i gotta see thos number that OPB got from more people. :D I'll be looking forward to your results!

[greenmaji]

OPB bins them, unless you know him, just look for the results from people buying from tankguys :D

[RangerXLT8]

Some people on XS say that @ high voltage on air, the prescotts they own run COOLER than Conroe.

Conroe uses alot of power, and creates alot of heat when the voltage goes up.

;)


Any processor, when run beyond spec is going to generate alot of heat.

Compared to Prescott\Presler, Conroe uses substacially less amount of the power. Presler XE TDP is 135watts, while Conroe XE is 95w. All the other Conroe\Allendale are rated for 65w tDP. Now throw in some overclocking and those numbers go up, but keep the same ratio.

Our resident mad scientest proth, did a neat little experiment with his Presler XE. He found that @ 5ghz, the 955XE draws like 280w( or something in that ball park). You will never find a Conroe that draws that kinda power:mad:

@ JC, I can't wait to see your proc frozen!

[Joe Camel]

too busy to be "sick" today :-/ think ill head out the door @ 3:30 though.

i love to see TankGuys posting :D

(sorry i couldnt hold out to pre-order through you...but if the next batch of chips OC better, ill sell this and order through you. you CAN get 6800's if requested, just didnt want to stock them right?)

[NoL]

Now you just need a second phase unit for other CF gpu. :)

[Joe Camel]

and the $ to pay you... oops i mean who ever could build such a crossfire master card SS unit ;) ;)

[Ross]

proth, now THATS dedicationActually, that might have been a faster way to do it and a lot more secure. If you can find mobo stand-offs that are ~3/16-1/4" taller, it would probably even go on a mobo tray like that. Nice idea proth!

[xgman]

a friend of mines compare with software vs probe to IHS:

[party animal]

does that say 83*C? Jeez, that is one hot cpu :mad:

[Joe Camel]

ahhhhh

this is like saying:

the thermometer in my car doesn't work, it reads 500
so i taped a thermometer on the hood, it reads 150

what temp is my engine?


:bang head

[Gautam]

JC, the core is soldered to the IHS...just how much of a temp difference do you think there possibly can be?

The engine is not fused to the hood, but the core is to the IHS. ;)

[party animal]

I trust the thermometer more than software, but what motherboard is that? JC, want to try software compared with thermometer so see if your Conroe really is 51*C load?

[Joe Camel]

JC, the core is soldered to the IHS...just how much of a temp difference do you think there possibly can be?

The engine is not fused to the hood, but the core is to the IHS. ;)
cant be that much better than the deltaT we get from HSF to core.
(or id assume we'd see more HSF's soldered to cores)

and then throw in the fact that the probe is on the side of the IHS; as far away from the core as you can get on the IHS.


i KNOW there is no easy answer... all we can do is look for comparative changes in the readings we each have (software and hardware differ between rigs) but with that in mind, ANY reading will let you compare load to idle.


sore subject (to me) sorry :(


EDIT:

in all honesty, i couldn't care less what the temp is really... if its stable and running benchmarks it must be "OK" and thats where i stop (these days)

[Alchemy1]

Those 3-4-4-8 Team group sticks are all sold out now :(

I don't think Team is making any more, either. They are being replaced by the 3-3-3-8 sticks instead. They are more expensive, $247 or so, but that's still a nice price for such good memory.

Sorry we couldn't keep the other stuff around longer. I think we sold a few hundred sets... it moved quick!

No, no, no....I was just going to pick up a set so I can start accumulating parts for my new build. You say the new 3-3-3-8 stuff is going to be $247, isn't the DDR21000 stuff that much ($250). Are the two sets of RAM similar (one runs tighter timings at a lower speed and the other looser timings at a higher speed)?

[xgman]

Just posted the pic (of an badaxe reading by the way) to show the really high temps may be off and not to worry about them. It's probably somewhere in between, but as someone earlier today said "Bottom line, if its stable, its cool enough. Go by how warm your block gets."

[Joe Camel]

ok that logic works for me :)

wasn't picking on you XGman... hope it didnt come off that way!!

[xgman]

One thing I like about this board is that if you go too high on the overclock, you usually don't have to clear cmos, although sometimes that message comes up for no apparent reason.

[Joe Camel]

SS just got turned ON :D:D:D

cool down and boot is all thats left

[Gautam]

:drool: :drool:

You might be right on the IHS...dunno...but if it were so bad then why does popping the spreading on AMD's usually only help by about 5C? (Especially when their mounting is clearly inferior to Intel's) :shrug:

You are the ghetto god of cooling though, I'm just a hack it together and switch it on sorta guy. :D

[Joe Camel]

LOL that description sound about the same :p
(i just take more pictures and use more DUCT-TAPE :clap: )

ok i can see the logic behind the AMD IHS removal only gaining a few C ... huh...

had a false start on the SS :( 5 min countdoun.... :crossfingers:

[xgman]

LOL that description sound about the same :p
(i just take more pictures and use more DUCT-TAPE :clap: )

ok i can see the logic behind the AMD IHS removal only gaining a few C ... huh...

had a false start on the SS :( 5 min countdoun.... :crossfingers:


just don't blow anything up. :mad:

[Joe Camel]

Good luck on the SS. Here's hoping for >4.4GHz :beer:

you'll GET IT!!!!!!


lets put it this way, im giving 32M a run @ 4.5 now :beer:



this is still @ the same voltage as i was running @ on water

cant find a temp program that can read <0 so im flying blind again ;) ;)
getting a -40C readout on my chilly1 SS , 1x Pi so "1/2" load , room temps are 75.....


DAMN LOCKED on loop #8 of 32M ....


lets try...??? MORE VOLTS :beer:

[Gautam]

YESSS!!!!!!!!!

Ohh man, where are all the XE/step 6 haters now???

Can't wait till I get mine! :D

[toddm27]

:eek: very nice, now you really have me:drool:

[Joe Camel]

hang on... im shooting for a :drool: Pi 1M for the first screen shot ;)

[Gautam]

11.4x is my guess, but please impress me further! :D

[party animal]

Oh boy! This is going to be great!

[Ross]

Nice gain on the SS JC! I am still not loving the Step 6...not sure how it can suck that bad on air/water. The jury will remain "out" until I see what it can do under the SS ;)

It better be more like 11.2 since I can do 11.7 @ 4.25 with 1:1 on 8000ULs.

[Gautam]

The speed needed to lower Pi times literally starts getting asymptotic. :-/

Usually 4.5-4.6 GHz times are in the mid/low 11 sec range. Though you are one of those that have a certain talent for Pi. :D

[dwschoon]

Man, you people have way too much $$ to burn. :) I am anxiously awaiting more results!

[Burdman27911]

If they really have too much cash to burn, then I'll gladly have half and we can (litterally) burn the rest... what a poor attempt at humor on my part :shrug:

It's looking very good again, I'm waiting 5 months more anyway before I start my build, so I'll have plenty of time for things to be worked out. Heh, I'll be pretty happy with an E6600 @ 3.2-3.4ghz.

Best of luck Joe!

[Joe Camel]

11.4x is my guess, but please impress me further! :D



you'll GET IT!!!!!!






http://img64.imageshack.us/img64/899/112pi1maj6.jpg






Chilly1 reading -42C





http://img208.imageshack.us/img208/4725/pict0262wn1.jpg







for the first 5 or so boots BIOS was giving me OVER TEMP warnings but then it stopped.... and then after a hard reboot they started....and then stopped :confused: but anyway, BIOS does seem to read <0 temps


EDIT:

think ill be picking up some D-ice tomorrow morning and "smoke" the Camel Pipe (http://www.ocforums.com/showpost.php?p=4551421&postcount=140) all weekend ;) ;)



.

[dwschoon]

Wow! It sure looks like its the beginning of very exciting times in the computer market!

[Ross]

The speed needed to lower Pi times literally starts getting asymptotic.

Usually 4.5-4.6 GHz times are in the mid/low 11 sec range. Though you are one of those that have a certain talent for Pi.:D I think SPi time per CPU clock change is pretty consistent...it is for me at least. That's not to say tweaks, NB/CPU/mem efficiencies don't come into play, but all else being equal, every "x" FSB delta pretty consistently yields "y" time delta for me.

I am almost done messing with this DQ6. I just dropped the Presler/7600GS in to test it out, but I'll be back on the P5W and SS tomorrow. P5W *needs* a BIOS that let's me get some Vc to the 6600 because the 6600 is a sick clocker...just can't do anything with it in the P5W :( 4.15GHz on water on the DQ6 (460FSB) @ just 1.55V :(

JC: Sweet ;) Thanks for getting that 11.2 for me because I called it, LOL. IIRC, the BIOS will go from -27.9 to +127C as it cools off (ie. -28C or colder screws it). ~1.65Vc?

EDIT: I must be a mind reader, I didn't even see your comment about the overtemp warning before I wrote that. In any event, if you often boot with a BIOS temp reading < -28C, you can switch the CPU temp to "Ignore" and that will keep it from hanging on the CPU temp at boot.

[Gautam]

Holy crap to both JC on that insane Pi result and Ross on that pyscho watercooled clock. :eek:

I don't understand why the E6600's seem to be THE BEST overclockers. I'm beginning to feel its got something to do with multiplier or something. It just doesn't make sense how they seem to kill everything else. Then again, they quickly lose their edge subzero quite obviously. Never in the past though has it been like this with Intel. You had to buy the best to get decent clocks, at least I had to. :-/ 670 vs 660, 955 vs 940, etc...In any case you definitely got two amazingly clocking procs there. I bet I could buy a lot of 5 XE's and not match em with my luck. :-/

[Ross]

I dunno, but I am having second thoughts about getting rid of it. I wasn't sure I screenied it, but I did. Water, 1.55Vc:

http://valid.x86-secret.com/cache/106972.png

Since I hit 475FSB on the board today with the 6700 @ 8x after making some changes, maybe I'll toss the 6600 back in and see where it suicides with the board no longer being an issue...or at least less of one. It's still pretty PO'd, but I am not sure if the BIOS memory issue is what's causing it or not. If the mem and the reboot issue can be fixed with a BIOS update, I'll probably keep the board, but I dunno how often Gigabyte updates :shrug:

Sorry to jack your thread here JC, we're all living vicariously through you. We all want to see <11.0x by the end of the weekend...hurry up on the DI ;)

[Gautam]

I'm sure he'll deliver. My step 5 was one out of 10 in country, guess it was the worst of the bunch or something. :-/

The fact that my Presler clocks roughly the same as your Conroe with similar volts and cooling should say somethng about my luck with intel...:rolleyes:

[Jame Gumb]

The "Camel Pipe" ROFL.... stuffed with that turkish blend of dice!

[Ross]

I'm sure he'll deliver. My step 5 was one out of 10 in country, guess it was the worst of the bunch or something.Me too. Personally, I'd like to see him do it on the SS and then about 10.4-10.6 on DI ;) Hehe, yeah, you haven't had a bunch of luck. I'll be REAL curious to see how the B2 6600s stack up.

[Qubix]

JC, good to see your X6800 is doing well for you so far, lets hope it has more to give on DI.

You guys are filling me with confusion! Personally I will be very happy if my E6600 does 3.6ghz stable 24/7/365 and upset if it does 3.4 or less and bouncing off walls if it hits 4ghz+. But after listening to others in this thread makes me feel I am asking too much from the E6600 for stability and then seeing Ross makes me feel like I am on the right track.

[Ross]

Don't take mine to mean anything. I haven't done any stability testing at all ;) 3.6GHz should be a no-brainer though and that should happen on air without insane Vc. I guess we'll need to see how the retail 6600s stack up, but the ES 6600s are better clockers than any of the others without a doubt.

[Joe Camel]

ok tried a little different stability test (05) and have come down into the 4.3's to get through a FULL (CPU too) test so that screen shot was Pi 1M stable ONLY!!!!!

dont think your going to "Game" @ these speeds!!

[orionlion82]

ok tried a little different stability test (05) and have come down into the 4.3's to get through a FULL (CPU too) test so that screen shot was Pi 1M stable ONLY!!!!!

dont think your going to "Game" @ these speeds!!

poopy.
only 4.3 stable on phaze?
barely (not really) stable on phaze?
at the same or higher Vcore than im running now on 90Nm?
and 3.5 ish on water?

WHAT?!?!?!
you broke my bubble.
this cant be.

[xgman]

it may be the board.

[orionlion82]

it may be the board.
i hope so. because if not, all the press kits had "miracle chips" in them.
and the boards crap out and blow up(at least the early intel's)and they take stupid vcore to get to clocks that arent really as high as... at least i had hoped, at temperatures that are pretty shabby.

my wildest dreams were about four and a half plus rock solid stable, with low volts and low temps, and about four ish with the same criteria for low end water/high end air. and realistically, i was thinking high threes for air /and four plus on water (rock solid stable)

this makes me want to upgrade my 478 rig.

[Gautam]

Did they really? Last I checked Pedro Rocha's proc got to 4.6GHz flat on his single stage for 1M pi...which I'm sure was HEAVILY tuned.

Your expectations were unrealistic for even the best ESes out there. I don't think anyone pulled off 4 24/7 stable on air, or even anything close. NGMA isn't netburst.

[orionlion82]

Did they really? Last I checked Pedro Rocha's proc got to 4.6GHz flat on his single stage for 1M pi...which I'm sure was HEAVILY tuned.

Your expectations were unrealistic for even the best ESes out there. I don't think anyone pulled off 4 24/7 stable on air, or even anything close. NGMA isn't netburst.

guilty as charged.
i believed the hype. -because this time it was different, and it was worthy of all the hype-
reality has been trickling in and trickling in, and last few days, its been pouring in... and its not as appealing anymore.

and sure i dont keep up with this stuff more than every few days, but things change so fast over such a steep gradient even in a few weeks time -when you do it that way- that it seems like ive suddenly been dunked into icewater. the slack between reality and hype has been taken up so fast to me it seems. not that im crapping on joecamels OC, much the opposite - i applaud him. - but im just not anywhere near as excited about conroe anymore as i once was.
now, sure its still wider and clock for clock gets alot more done, but im not sure im liking the new rules for this new game called conroe.
and i thought my expectations were merely on the high end of reasonable, and not outlandishly insane...
hmmph.

[dwschoon]

guilty as charged.
i believed the hype. -because this time it was different, and it was worthy of all the hype-
reality has been trickling in and trickling in, and last few days, its been pouring in... and its not as appealing anymore.

and sure i dont keep up with this stuff more than every few days, but things change so fast over such a steep gradient even in a few weeks time -when you do it that way- that it seems like ive suddenly been dunked into icewater. the slack between reality and hype has been taken up so fast to me it seems. not that im crapping on joecamels OC, much the opposite - i applaud him. - but im just not anywhere near as excited about conroe anymore as i once was.
now, sure its still wider and clock for clock gets alot more done, but im not sure im liking the new rules for this new game called conroe.
and i thought my expectations were merely on the high end of reasonable, and not outlandishly insane...
hmmph.

I understand your concern, but I think there is still plenty to be excited about with conroe. The price/performace factor is still awesome, even if it tops out at 3.5 gig stable on air. Not to mention, the fact that it will drive all the other cpu's down in price. The first revisions of any cpu usually dont clock as well as the later ones. Im willing to bet that later revisions will do much better than these first ones.

[mikeguava]

Awesome 4.5GHZ on climbing!!! Keep it going JC! Eventually only Vcore will get you more clocks - you might be getting closer to this point now

EDIT: I have mentioned this before - but with some Intels - the max clocks appear to be impacted - positively or negatively - by the core clock multiplier. Not that it fully makes sense to me - but I agree that is worthwhile experienting with it...


YESSS!!!!!!!!!

Ohh man, where are all the XE/step 6 haters now???

Can't wait till I get mine! :D :o right here...lol

[Joe Camel]

The "Camel Pipe" ROFL.... stuffed with that turkish blend of dice!
:beer:

my ExtremeSystem post:

http://xtremesystems.org/forums/showpost.php?p=1519703&postcount=12


as much as i'd like to break into the 10's (Pi 1M) with the SS, i want D-ice results B4 the true release date. (dont you ;);) )


:beer:

[party animal]

of course we do! Keep up the good work JC!

[Joe Camel]

25lbs of D-ice :thup:

:D

have to spend the day (after work) with the wife, but "after dark"... its ALL ICE baby :beer:

ill post up ASAP :)

[Gautam]

Working weekends sucks! 5 days is enough for me...Yes I think Mike and I might be onto something...give the 9x a shot for S's and G's...no guarantees of course, this is just going off how amazing the 6600's seem to be.

[Joe Camel]

problem with 9x is i cant seem to get the mobo to get past about 442 FSB (even with the 8x) so that limits my total OC (and lowers Pi times)

but with 25lbs of D-Ice, i think i can try a "few" things out ;) ;)

[Gautam]

Yeah I wasn't thinking clearly. Down the road you could MCH mod it and drop some chunks of DI on the NB heatsink...insulate of course. Too much for a day though I think. The multipliers would be best to investigate on air/water to first figure out if its even worth it or not I guess.

[Joe Camel]

.


booo-yaa :D


http://img204.imageshack.us/img204/8162/10703pi4744vo7.jpg



FULL size link:
http://img204.imageshack.us/img204/9229/10703pi4744bc4.jpg




the rig:


http://img53.imageshack.us/img53/5306/pict0263ml3.jpg

[greenmaji]

SICK SCORE JOE!!!!:attn: :attn:

[Gautam]

And there goes the notion that the step 6's can't keep up, dead as a doornail.

You never cease to amaze, remember, post in hwbot, please!!

[Joe Camel]

trying tighter timings then ill slow the mem and go for some 3-2-2-1 :D

going to be a FUN night :beer:

[d94]

dayummm...

[Jame Gumb]

[paris] [/that's hot]

[pscout]

Nice work Joe :thup:

You make this seem so easy.

I guess i will have to wait till a 6600 arrives before i see OC'ed folding benchies ... 25 lbs of DI just ain't enuf. :D

[dylskee]

[paris] [/that's hot]
LMAO!!!!!!!!!!! I love that girl!!!

Nice job Joe, very impressive to say the least.

[Joe Camel]

Nice work Joe :thup:

You make this seem so easy.

I guess i will have to wait till a 6600 arrives before i see OC'ed folding benchies ... 25 lbs of DI just ain't enuf. :D
hard-ware and tools... just like ANY job :)


thanks guys :beer:



PS Pi 1M stable and FOLDing stable are WORLDS apart!!

[mikeguava]

AWESOME clocks and times - keep em rolling

And there goes the notion that the step 6's can't keep up, dead as a doornail.


Told ya

[IYIENACE]

Woah man that's screamin! I agree some of you make it seem like pouring dry ice on your mobo is some how sane lol. :beer:

[Stilletto]

Geez Joe....that makes yours 300% faster than my max oc 805.

That's just insanely fast. What is the current 1m Pi record anyhow?

[proth]

schweet!! blew right by 4.5. Very nice

[El<(')>Maxi]

Nice 1M Joe, DI made a little difference ;) CF soon?

[Joe Camel]

OMG crossfire :bang head

im in no major rush to add that much headache to my future ;)

after this 3D competition, i need a break ;)

i just really wanted to get y'all a look @ the Retail 6800's B4 ya blew BIG $$











----------










Retail (Newegg) 6800 gets SMOKED (good way)









title explained: i named the D-Ice tube
The Camel Pipe
*- Just SMOKE It! -* (http://xtremesystems.org/forums/showpost.php?p=1519703&postcount=12)




Retail 6800 from Newegg:


http://img477.imageshack.us/img477/1254/pict0260hy8.th.jpg (http://img477.imageshack.us/my.php?image=pict0260hy8.jpg)




#3 Pi 1M (mbot)



http://img204.imageshack.us/img204/8162/10703pi4744vo7.jpg







#2 Pi 32M (mbot)




http://img205.imageshack.us/img205/9382/32m1149188im2.jpg



6800EX
ASUS P5W-DH (un modded but improved cooling)
2x gig Corsair XMS 6400 CAS3
Zippy 700w PSU
D-Ice Cooling



http://img53.imageshack.us/img53/5306/pict0263ml3.jpg



i know: it reads low

[Ross]

Nice job Joe! I wish it did a hair better on water, but nice to see they're not total bunk ;)

[Joe Camel]

.



took #1 in 32M


:beer:

OCF #1 world mbot Pi 32M baby :attn:




http://img80.imageshack.us/img80/7137/32m112850xc1.jpg



http://img208.imageshack.us/img208/6624/1worldpi32ce0.jpg

[pscout]

very :cool:

:clap: Congratz on another #1 :clap:

:beer: :beer:

I wonder what else i will see in the morning :D
/edit and I saw your posted #1 :thup:

[party animal]

Nice job JC!!! Very very nice. Can't wait to see the 3D numbers :D

[xgman]

Congrats!!

[Gautam]

Indeed, once JC gets to 3d it's gonna be a SLAUGHTER.

[orionlion82]

congrats.

Well I guess it would be nice
If I could touch your conroe
I know not everybody
Has got a conroe like you
But I've got to think twice
Before I give my cash away
And I know all the games we play
Because I play them too

Oh but I
Need some time off from your pipe smokin
Time to pick my clocks up off the floor
And when that rig comes down
Without high voltage
Well it takes a strong man JC
But youre showing us the way

'Cause I gotta have phaze...
Mmm, I gotta have phaze
'Cause I gotta have phaze, phaze,
Mm 'cause I gotta have phaze-a-phaze-a-phaze

gawd that was so fun :santa:

and i just built a new cooling system for conroe.
maybe if i end up with one and not get kentsfield, ill see if i can find a way to get water further...:beer:

[fritzman]

Wicked work JC... Congrat's man.

You are a LEGEND!

[Gautam]

Reviving this thread cause I got one too! Followed tons of JC's advice and insulated exactly as he showed earlier in this thread. Barely missed his SS speeds by 20MHz but that's more than acceptable for me considering I'm "just" on a Mach II. This is probably the first Intel proc I've oced that doesn't make me feel like crap.

11.297s 1M Pi @ 4510MHz
http://mysite.verizon.net/gautamb/1m.JPG

11m 43 s 32M Pi @ 4455MHz
http://mysite.verizon.net/gautamb/32m.JPG

And here's one you don't see often, Prime95 for 10 hours and 44 minutes at 4206MHz :)
http://mysite.verizon.net/gautamb/prime.JPG

I think 4.3GHz daily may not be a stretch with this proc.

[Silver]

Congrats to both of you. Looking good. Very, very fast processors.:clap:

[Ecca]

you guys are amazing

[Jame Gumb]

Nice friggin oc. What board was this on?

[OnDborder]

Holy Moly,, way to go guys..:attn:

[Gautam]

Thanks guys...JC and I spent a lot of time bouncing ideas off each others heads. :p But mostly its the procs clocking so well and the subzero cooling. This one is EASILY 300MHz faster than my E6700, closer to 350MHz actually. It cost basically everything I have, but was worth it.

We're both using P5W DH Deluxes. Now thanks to the unlocked multiplier, I've found that my FSB tops out at 415MHz. :-/ This could be a problem for lower-end chips. My inkling is that Conroes should usually stick to this board anyways, and Allendales would be best off with the Gigabyte offerings, or possibly the P5B Deluxe.

[Jame Gumb]

I don't think most people realize the skill it takes to oc where you guys do. A lot of tips and tricks, from software to cooling. I know Coolaler, Andre etc aren't going to give away all their tricks. Nice to see you guys sharing ideas. :thup: Allendale w/ Giga's, BioStar & P5B-D w/Bios 0507 are stomping now. Still a lot more BIOSes to go.

[pscout]

Nice work Gautam! :thup:

:beer: :beer:

I think the p5B deluxe is the best of my bunch atm ... at least with a 6400 ... beats the ds3 by 11.5% folding!

[Sucka]

Nice results. I can't say i'm very pleased with that prime shot....

;)

[Joe Camel]

Nice results. I can't say i'm very pleased with that prime shot....

;)
.

LOL

sadly Sucka, EVERYBODY seems to want to know "dual-prime - 24 hour - super ultra stable clocks" these days :rolleyes: ;)



sure it can do Pi 1M in 5 sec (so what) is it PRIME stable??!!

:mad: :shrug:


.

[Sucka]

sadly Sucka, EVERYBODY seems to want to know "dual-prime - 24 hour - super ultra stable clocks" these days :rolleyes: ;)

Unfortunately they've wanted to know this long ago as well...

:-/

[MadMan007]

Thanks guys...JC and I spent a lot of time bouncing ideas off each others heads. :p But mostly its the procs clocking so well and the subzero cooling. This one is EASILY 300MHz faster than my E6700, closer to 350MHz actually. It cost basically everything I have, but was worth it.


Does this tell us something about the speedbinning of these CPUs? An e6700 with a 10x multi should still be fsb limited - 4GHz+ with a 10x multi - so if the x6800 regularlry goes higher I'm thinking the speedbinning is pretty strict on these CPUs.

[greenmaji]

Does this tell us something about the speedbinning of these CPUs? An e6700 with a 10x multi should still be fsb limited - 4GHz+ with a 10x multi - so if the x6800 regularlry goes higher I'm thinking the speedbinning is pretty strict on these CPUs.

speedbinning :p
http://www.xtremesystems.org/forums/showthread.php?t=110808
http://www.xtremesystems.org/forums/showthread.php?t=111081 <-retail LOL

J/K

[Epox4life]

These X6800s under subzero cooling are killer clockers.

[Ross]

Nice clocks Gautam...nice that you finally got a solid one. Did you try it on air/water at all and what would your Vc be for that "daily" 4.2GHz clock?

[Gautam]

JG- thanks for that, you are absolutely correct in that a lot of effort goes into this. Trying to keep up with Ross and JC for example on similar setups gave me an even greater respect for them. They and many others definitely work very very hard on their overclocks, some very very fine tuned results that take a helluva lot more both on the software and hardware side than just hitting a few buttons in BIOS. But you won't ever see any secrets being kept here, at least I should hope not. I teach everything I can but always have lots to learn. I can't help but quote Skip-

We are a team. We are a community. We are a fellowship made strong by mutual respect and shared dedication to the task of enriching all who come here. We are respectable, for we share our bounty.

No ones out to withold any info here, we're all here to help one another.

Madman- It's too early to fully tell what's going on, but it seems like retail 6700's can be expected to fall around 200MHz short of X6800's barring FSB limitations. Once again though, for 24/7 usage, I wouldn't really expect the 6700's to be FSB bound most of the time. On subzero cooling for benchmarks, yes, very possible and likely. On air, I still think a 6600 at about 9x400 is what's to be expected and optimal.

One random thought that I will leave you guys to chew on is that JC's OEM X6800 clocked nearly 100MHz lower than his retail...other OEM results also haven't be all that, but the few retails I've seen seem to be very competitive against ESes...just a thought for prospective buyers based on my own observations. Take it for what its worth.

Ross- vc for the pi runs was 1.675 BIOS set and 1.65v for the Prime run. I went straight to the Prommy, sorry. Couldn't help myself. And yes it is about time that big blue treats me right for the amount $$$ they've gotten from me this time. :p I still can't get over the fact that this proc actually clocks better than my Presler. :D I don't know if that's sad or what.

Sucka- lool...that's why I deliver on all counts, no complaining possible about my overclocks ;), I cater to both the die hard benchers and the prime nazis. :D

[MadMan007]

Fellowship of the OC!

Ok, that was pretty nerdy :p

Yea not enough samples yet to really know - except maybe if those lucky people (dealers) who can sift through numerous chips kept detailed records - it seems as if NA users are getting more of the low-end chips which end up FSB limited. Heck even the weekly mobo updates can change the scene a lot. It's interesting to read about, I'll be happy for the guinea pigs when I do mine in a while ;)