hmm...the easy solution to this would be to just start a new thread, where it can be discussed without worrying about ruining another threads theme. Look, I'll do it right now.
Printable View
hmm...the easy solution to this would be to just start a new thread, where it can be discussed without worrying about ruining another threads theme. Look, I'll do it right now.
phi|os there already is one
This is basically what we're all trying to say..Quote:
Originally Posted by Absolute_0
Oops. lol.Quote:
Originally Posted by itznfb
Where?
Actually the last thread is his
Ok, so getting back to what this thread is really about, I'd like to present to you guys the very strangness that is my system.
I've been trying to get my system stable at 2.9ghz with 1.45V. About ten failed 30min prime test later, I find out that it is super stable at the above setting when I further decrease my RAM speed. I'm using OCZ DDR500 2x1GB gold RAM. My RAM was at 240ghz or close to it when it wouldn't pass 30mins prime. I used a lower divider, my RAM is now at 223, and it's been priming for like 4hrs now...
Does my 165 CCBBE 0610 like lower DDR? Anyone else run into a situation like this? My hardware is in my sig if you have a question about it.
A64 tweaker ss please, we can probably get your settings tweaked up and stable :)
lol...this thread became such a joke now.
seriously, i think your in the wrong thread...this one just so happens to be about Opteron overclocking. i suggest you go back to school and learn to read. it just *might* help you figure out where you should be posting.
i actually have that problem as well. my ram is 233mhz and i had to run it at 168mhz to get it stable at 335HTT.Quote:
Originally Posted by phi|os
Yes, I think the plethora of "stablity" threads should satisfy even the most avid thread FUD'der!Quote:
Originally Posted by itznfb
*Directed toward Repo*----> I'll get some screens up ASAP.
Ok. Here it is. Two hours prime actually. But it wouldn't go over 30mins until I dropped to lower mem divider...
http://i75.photobucket.com/albums/i3...3/fa6d0cb4.jpg
Its probably your memory controller and the DRAM settings you're using. You may need to adjust the drive strength, tref, max async latency, ect. Try 7 and 3 for the drive strengths, 3684 for the tref, 8.0ns and 5.5ns for the latency and enable Dynamic counter as well.Quote:
Originally Posted by phi|os
I agree. To those who are still on this, the truth is, you don't know why prime fails. If blaming it on heat makes you feel better than that's fine. To each his own.;)Quote:
Originally Posted by Repoman
EDIT: Whoa, I'm late.
Hate to say this dude, but your not priming right. If you look at your post you will see that you have assigned two instances of prime to only one core! You should be running one on core 0 and the other core 1. Also you want to select the same torture test i.e small FFT for both instances of prime.Quote:
Originally Posted by phi|os
Hope this helps.
No affinity. That means the two programs get distributed between both cores...
Small fft's seem to be too easy of a test. Larger fft's make memory and cpu work harder, and produce more heat. Maybe it's just a preference. Task manager shows both cores going at it, and no affinity means just that, no affinity. thanks for the help though.
Still stock air w/ ceramique, undervolted... pretty sweet core, dice this core will see come friday ;)
Rosetta for about.... 4-5 hours.
http://img69.imageshack.us/img69/190...eron0tl.th.jpg
If you dont set the affinity then you wont really know which core is failing when it errors out. Small FFTs are the hardest on the CPU because they fit into the L2 cache and stress the floating-point unit and cache the most. Large FFTs stress more of the RAM, memory controller, HT link, and CPU combined.Quote:
Originally Posted by phi|os
Well things are getting interesting for me. It wouldnt prime last night for more than 5 seconds with 1.52V @ 2.9Ghz now I after 2 burn in cycles it passes dual prime for over 2 hours with 1.46V @ 2.9Ghz. Oh, this is a CCB1E 0606VPMW that I got from Monarch on monday for those of you lost in the stability discussion.
Set priority higher babalouj, fails faster... :)
Nice clocks
What stepping Eboy0? Btw, I've tried both 9 and 1 for priority, both ran just the same.
It's one of my 3 CCBWE 0550UMPW's all low 150 batchs. All going under ice :) seems like a pretty sweet core, this one in particular runs the ballistix @ 4 tras unlike the others which run them @ 3, no diffrence really..
Hi all!
Just wanted to let everyone know about my new 3GHZ opty 165...I am proud owner and cant wait for winter to come back and probably hi 3100mhz! Thanks meatfrog for selling it to me. :-)
Regards
I actually had it at the highest priority, I just turned it down to post the screen shot. 3Ghz, here I come! Oh and its still priming happily away approaching 4.5 hours.Quote:
Originally Posted by eBoy0
i got good news and bad news:
good news: i got my CCBBE 0610DPMW today :D
bad news: i got it from the US and the stupid azz seller didnt mark it as gift so i had to pay $60 in customs fees to pick it up!!!! :mad:
anyways, gonna run some quick tests for screenshots on the old cpu then pop this sucker in. it better OC like hell to make up for the trouble its given me so far.
The priority is locked, so you used task manager? Btw, you can use your computer at priority 9.Quote:
Originally Posted by babalouj
You should uncheck no affinity for maximum stress. You want each core to correlate to its own instance of Prime 95 so that the work load between them is uniform and even.Quote:
Originally Posted by phi|os
I get about the same heat and fail times no matter what priority and small/large FFT i use. I generally run small.
@ babalouj
Burn in and lower temps ;)
With my 0610DPMW in sig, the first day i got it, 3 Ghz with 1.52v failed instantly. I mean less than a second. The next day my CPU had been burned in overnight, and my room was a little cooler, i tried 3 Ghz with 1.48v and it kept going... stable with less voltage the next day. Temps & i am pretty sure burn in.
What's the WaferID?? Can you post the serial #? You can leave out the last digit if you like.:p:Quote:
Originally Posted by blazin-asian
I'll try it, but from what I remember when I started using prime, small fft's were not as stressing on my system. But I'll set the affinity..so I can see which is weaker...so I can tell you guys about it...so that I can just be all :banana::banana::banana::banana::banana:y that core 0 is weaker than core 1...:stick:
IMO, small fft's suck.
wafer id is: 1434904C6035xQuote:
Originally Posted by thunderstruck!
at least the seller paid the extra $15 for 3 year monarch warranty if anything happens. then again, im sure if i were to RMA it i would most likely get a 0551UPMW. :slapass:
gonna pop it in in a few hours. also going to try a BT mod for better temps since i have to take the mb out anyways. the only thing i hate about the BT :stick:
Yeah, I used task manager and set them to real time when I am not using my computer but I cant do anything if I try to surf the internet while dual priming with real time priority.Quote:
Originally Posted by thunderstruck!
Yeah, burn in is the real deal but I didnt think it would make this much difference. I cant wait until after my move to push the limits further!Quote:
Originally Posted by Absolute_0
LOL, i think i know who you are. You bought the chip from me :p. I marked it as merchandise because that was what it was and i dont believe you ever requested it be marked as gift. Also, it wouldn't matter if its marked as gift or merchandise, you still pay the custom duty dependent on what the value is marked as. I marked it as only $300 even though it was actually $330. If i had marked it at a low value then i would not of been able to insure it at USPS. You can only insure for the declared value. If it had been lost in transit then one of us would of been SOL. I'm going to check into third party insurance agencies though for the future in case a seller wants me to LIE on the custom form and mark it as a gift with a low value lol.Quote:
Originally Posted by blazin-asian
Is it true that CCB1E's run hotter than CCBWE's? I used to have a CCBWE opty 0607 that i sold and temps at load stock would only hit 34c on a scythe ninja, but when i bought a new opty CCB1E 0608MPMW my load temps were 45c. I remounted the heatsink 4 times and still same results. My water cooling though will be all together by early june, but i'm still debaiting or not to take off the ihs because i really think i have a defective ihs.
^Yep, sounds about right
I hit 45-46C at 1.45v on water with this CCB1E 0608, until I cut the IHS off and now it loads at 34C, same voltage :)
edit: 38C naked with ceramique, only got it to 34C with liquid metal paste
Wow huge improvement man, thanks for the reply. With results that much improved i will definantly remove the ihs. How does that metal paste work, i've heard somewhere that it can leak from under the heatsink. I want to try that stuff out because i heard it works better than as5, but the whole leakage thing worries me.
ccb1e 0609fpmw is ready to go! :)
here's a couple pics i took while swapping cpu's for my fellow geeks to check out
couple pics of the motherboard from hell
Wowa! the opty all the way to the left is 005 :slobber:Quote:
Originally Posted by itznfb
henyways, lemme go clean up the drool, in the mean time answer this.
Why are you running single channel?
You have all these processors, yet you're stuck with this mobo...lol?Quote:
Originally Posted by itznfb
I got a A8R32-MVP lying around if you want it for a...um, well, pm me if you're interested.
hahaha, thats funny you ask that... notice how i mentioned the motherboard from hell? well, in msi's little fantasy world, they decided to put dual channel in the slots right next to each other. i guess they figured it would generate a little more heat... cuz thats goo... wait no thats bad. :slap: f'n msi.Quote:
Originally Posted by eBoy0
hah, you know? isn't that just retarded, i should have known better, i should have baught the dfi first, no i can't afford it and i'm stuck.Quote:
Originally Posted by phi|os
looks like you are gonna have to choose between the CCBBE and the CCB1E soon...freaking hilarious. I could only imagine being stuck with that board. It's hard enough for me to be patient with my own board, and it's supposed to be good.
Speaking of which, I had that problem where my board wouldn't run my RAM at its normal speed(DDR500) and OC right at the same time. I enabled the Dynamic Counter, and when I left for work it was priming away at 40mins, as opposed to the 8mins to farted at last time. Hopefully when I get off of work it will be chuggin away still.
I found this little tweak the other day and thought i would share it for those of you that might not of already knew how already. I like having the Windows Task Manager running in the system tray at all times so i can monitor the CPU load. I added a registry fix so it would automatically load at boot minimized in the system tray.
Add a REG_SZ value to the following registry key:
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\Curr entVersion\Run
Give the REG_SZ value the name TaskManager and then give it the following string:
"cmd.exe" /c "start" /min taskmgr.exe
very cool, i'm the same way and love to see how many times my processor will fart in a given moment:slapass:Quote:
Originally Posted by burningrave101
I had a feeling that was the case, reminds me of my old nf3 ultra-d :)Quote:
Originally Posted by itznfb
What Opteron 170s are shipping out now?
Any good ones? Or I just shouldn't get it?!?! =(
You bastards made me do it!
http://www.sspmustang.com/OT/165ftw.JPG
Yes no affinity balance things between the two cores, but affinity assures you you're fully stressing both cores.Quote:
Originally Posted by phi|os
And yes, LargeFFT seems the most logical test along with memtest. Some Blend to be sure the IMC works fine. My order is memtest test5 -> largeFFT -> Blend. Priority9 and affinity. And prime over SP2004, it starts (and fails) faster :D
I go the fast route, it is, try to boot at 334x9 high voltage then lower volts if it boots or lower HTT if it fails. It's much faster.
haha. also the reason why i "downgraded" to an opty 165 :)Quote:
Originally Posted by fhpchris
well, my CCB1E 0609FPMW is running about 5c hotter than my CCBBE's. i was able to get it to boot at 3100mhz though, which i was unable to do with the CCBBE's. here's the nut shot though. anything above 2600mhz it fails prime almost instantly. yet runs any other stress app fine. i think i'll try again tonight with a clean OS.
What does it clock like m8?Quote:
Originally Posted by blazin-asian
Im still waiting for my CCBBE 0610DPMW '904' to be delivered from the states. . .
Also what's the best utility to 'burn in' the cpu at the moment?
i've been using cpu burn-in. it works pretty wellQuote:
Originally Posted by scoobie dave
http://users.bigpond.net.au/cpuburn/
Ok here it goes:
I only wanted to show my results (had this result before but no valid screenshot posted) but then an idea to get a structured DB came up to me.
I think a screenshot to get in the database (primestable) on the first page of this thread should look like this:
http://img102.imageshack.us/img102/2...05db0ea.th.jpg
Data that has to be on it:
- 2x SP2004, with affinity set to one core for each instance and completed one round
- Task manager with the CPU-time column enabled showing the sp2004.exe process time
- 2x CPU-Z, one with the CPU and one with the memory tab
- Smartguardian (or another temp monitor)
I think everyone on this forum can judge for himself whether he finds it stable or not. It’s just that all the data is consistent and based on facts. Now, you can consider your CPU database-stable. What primestable or 100% stable is, can be discussed overhere: http://www.xtremesystems.org/forums/...d.php?t=100538
I’m pretty happy with my results now, but the CFX3200-DR will be here tomorrow/Thursday. I’m using one of the first SLI-DR’s non-expert at this time, so I hope the CFX will let this CPU scream even more.
Now show some more results! Any new 170’s kicking-ass?
BTW, pic of my (I may say golden) CPU: (Yep! IHS is still on!)
http://img126.imageshack.us/img126/8...41447ob.th.jpg
And my screen :)
http://img112.imageshack.us/img112/9...e3ha7op.th.jpg
btw does anybody know this wafer number of 0610?
143483AC60 ???
CCBBE0610DPMW
1434904C60XXX =These are the best ones.
143483AC60XXX=These are not so good. But Good.
My first post =)
Well anyways i recently bought an opteron 165 ccb1e 0609fpmw but i know where i can get my hands on a cheap ccbbe 0615dpmw. Which one do you guys think is better? or should i spend the extra cash 2 try to get a ccbe 0610dpmw?
The CCB1E 0609FPMW may overclock well and it may not. The 0610 overclocks the best on average right now. I havn't seen many 0615's to tell how well they do.
How's the 0609 performing then? :cool:Quote:
Originally Posted by ineedaname
i just ordered it haven't recieved the ccb1e 0609 yet so can't tell ya.
0615 vs 0609, i would definatly go with the 0615. the temps on my 0615 were incredible. 3022mhz stable on air.... thats the highest, or one of the highest 165's on air so far. i'm finding it pretty hard to get my 0609 stable @ 2900 due to high temps.
We will use the tried and true Ebay method of finding the good ones then ^^
It may be good, it may not. However it is retail, and I will get the first chance to have a go at it.
If I do not like it, its going to ebay :)
THX , Rufus, you ve helped me a lot :)Quote:
Originally Posted by Rufus7
Insane results.Quote:
Originally Posted by ChkDsk
I think Prime95 is ok too, and who cares about the memory tab? :p
It gives people an idea what dividers are working well with that stepping and also if the memory controller is strong enough to run those high CPU speeds with the RAM at a high speed as well. For example is probably going to be easier to run a CPU at 3GHz and the RAM at 200MHz then it is 3GHz and 280MHz.Quote:
Originally Posted by thunderstruck!
all CCBBE/CCB1E overclock roughly the same. The difference between the weeks is very small compared to the difference between CCB1E/CCBBE and CCBWE. Most of it is random chance as usual... although after having a spectacular 0608MPMW i'm partial to saying 0608s are particularly good ;)
I had to raise my RAM voltage from 2,8 to 2,9 to prime at 2900MHz and up. At first I thought my board didnt handle the high HTT, then I realised that I couldn't run superpi. With a little higher voltage at RAM it was no problem to run superpi and I could run prime at 3GHz, something that I couldn't do before.Quote:
Originally Posted by burningrave101
So it might be an idea to set the RAM to a low clock when testing the max cpu clock?
So I am assuming the CCBWE 0550s that do 3ghz on air/water are about as coldbugged as the CCB1Es and CCBBEs ?Quote:
Originally Posted by Absolute_0
i believe the CCBWE arent prone to coldbug as much as CCB1E and CCBBE.
Hey would u guys suggest the big typhoon or the zalman 9500 if i wanna get the best air overclock?
big typhoon with a 80-100cfm panoflo
Looks like Newegg has CCB1E 165s :)Quote:
Other Thoughts: Get one soon (this week). CCB1E
Reviewed By: CCB1E, 5/23/2006 5:06:32 PM
Normally of course not. I am saying, are the 0550s more coldbugged than, say 0513-0517s or 0530 CCBWEs.Quote:
i believe the CCBWE arent prone to coldbug as much as CCB1E and CCBBE.
No, they are all the same. Not coldbug until about -30C core temps, or maybe colder. When it comes to coldbug, same steppings act identical (unlike OCing, where week means everything)Quote:
Originally Posted by fhpchris
I have a Zalman CNPS9500 LED and it cools pretty good. It helps if you have good case airflow though. The spiral copper heatsink is great but the fan doesn't push a whole lot of air.Quote:
Originally Posted by ineedaname
Thats odd. I could swear I saw Dumo or Stephen saying that CCBWE fluxuates alot.Quote:
Originally Posted by thunderstruck!
I know that even CCB2E in the FX 60s can fluxuate alot( 190-250 HTT ish).
I just assumed that common sence would say that the gap between B1 BB B2, and something like BH would make BW a larger span....
I used the Zalman before I went to water, still use is on another machine. Great temps & quiet, but my dual core & single core temps were always pretty high when overclocking & priming (fan was on high). The TT will be a bit cooler and the amount of air it moves will cool other parts. If your looking for extreme cooling & don't care about noise, TT. The Zalman IMO is the best low noise cooler for overclocking which is why I picked it over theh TT.Quote:
Originally Posted by ineedaname
Quote:
Originally Posted by burningrave101
at 1.475v + the 9500 wasnt nearly enough for me temps went to high 55c+ and it really limited my oc. On my single core though it was perfect. i switched to the tunic tower 120 and temps dropped alot 10c but still felt it was a little high so i mounted a second120mm with som nylon fishing line :D and it dropped 4c more so now im happy. the 120mm fan is a big plus since its really quiet with its fan controller.
have decent airflow with 5x120 and 4x92mm pushing air in and out of my chieftech mesh fulltower and 2x120 + 2x92 pushing air in the right direction inside the case.
also heres a test of all three coolers (9500 , TT BT tunic tower) in a controlled
temperature scenario http://www.xbitlabs.com/articles/coo...ower120_8.html
sory about spelling and so on i just woke up :D
//Smurf
I currently own a 165, but it was never a great clocker 2.4 on stock volts.
I'm thinking of trying my luck with another chip, I wondered if I wouldn't be better going with a 170 this round.
Any suggestions?
Quote:
Originally Posted by Smalltimer
youre better of with a 165 since they seem to clock better than 170s.
try to find one of the god steppings on first page of the thread , perhaps flickerdown.com has somthing good in store :)
hiya compgeek , you get around, i have just got a ccb1e 609fpmw opteron 165 as my 175 will not go above 2820 at any settings or voltage, couple things if you could advise i am thinking of flipping the heatsink off and also as i would have to run at high fsb 330 + any other settings to look out for my ram will run up to ddr626 stable, so i would have to use a dividedr tested using a divider on my 175 and it did not like the 9/10 divider , any tips
thanks
Glenboy
[QUOTE=CompGeek]Lets keep the flaming to a zero this time guys, make the mods proud ;)
Maybe we can get a sticky? :) EDIT: Thanks IFMU
Known good steppings that are capable of 3ghz:
CCB1E 0550VPMW
CCBWE 0550UPMW (Some are crappy actually)
CCB1E 0605XPMW
CCB1E 0606VPMW
CCB1E 0608MPMW
CCB1E 0609FPAW
CCB1E 0609FPMW
CCBBE 0610DPMW
CCBBE 0615DPMW
Honorable Mention:
CCBWE 0530TPMW (2.9ghz avg.)
CCBWE 0546XPMW (2.85ghz avg.)
CCBWE 0609FPAW (Unconfirmed precise averages, but seems to be 2.8-2.9)
AVOID if you want to hit 2.8ghz+:
CCBWE 0534SPMW
CCBWE 0543TPMW
CCBWE 0546MPMW
CCBWE 0551 (Any packing code) (2.7-2.75ghz 1.55v)
CCBWE 0602 (Any packing code)
CCBWE 0607 (Any packing code) (2.65ghz 1.55v)
I currently own a CCBWE 0546XPMW.
Sadlly it will not put out anything over 2.65 @ 1.55v - I will try my hand at water cooling before passing it off it might just need better cooling :)
I am hoping for 2.8 :)
Does anyone know if there are places to purchase tested chips?
Hey smurf if you're so enthusiastic about the cooling why don't you try polishing the block on the tuniq tower? I'm sure that'll get you a few degrees cooler. Judging from the picture on the link u sent, it could use a good polishing.
BST threads in the forumsQuote:
Originally Posted by Smalltimer
I must be blind! I can't find anything like that except the online deals forum which states no selling etc.Quote:
Originally Posted by itznfb
was going to but when i recived mine it was a whole lot better then the tested cooler so didnt have to :)Quote:
Originally Posted by ineedaname
Quote:
Originally Posted by Smalltimer
100 posts then you will find what yore looking for i belive.
Thats awfull, then I must spam somehow to get 100+ posts, bad for me :mad:Quote:
Originally Posted by Getttosmurf
True, dave at flickerdown.com has 12 CCB1E 0609FPMW and 12 CCB1E 0606VPMW.Quote:
Originally Posted by Getttosmurf
dont think it will unlock if you spam ....Quote:
Originally Posted by Kashelz
Hehehe too much spam can = ban :slapass:Quote:
Originally Posted by Getttosmurf
Very interesting!Quote:
Originally Posted by mursaat
Is this a reputable seller? If so I will pick one up asap
i've purchased a few chips from him. i would recommend doing business with him.Quote:
Originally Posted by Smalltimer
The system should also take into account, user registration date, I have been a member on this forum for several years now. I guess I didn't have much to sayQuote:
Originally Posted by alpha0ne
i would recommend spending the extra 18 bucks and get a CCBBE 0615 from tankguys. its a much better chip than the 0609 or 0606.
Hey peeps, just wondering, i've got a chance to get an opteron 175 ccbwe 0543 tpmw. would this be better than a 165 ccb1e or ccbbe (one of the good steppings of course). will be putting whatever i get under water. thanks
There doesn't seem to be a choice in stepping /batch with tankguys though. The other seller is guaranteeing 0606 VPMW, they can do 3Ghz according to the list.Quote:
Originally Posted by itznfb
Unless I'm wrong
Er... nevermind, I see that now - Well now I have two choices, I emailed each to see if they will ship to Canada.
Whoever responds with a yes first will receive the order
um... what?Quote:
Originally Posted by buffarilla
are you asking if you should get a CCBWE 0543TPMW or a CCBBE?
already replied, mate....Canada is definitely doable. :)Quote:
Originally Posted by Smalltimer
dave
always.. :)Quote:
Originally Posted by Getttosmurf
CCB1E 0609 FPMWs (retail)
CCB1E 0606 VPMWs (tray)
:)
dave
Ah, Dave!
I just got my 0609 FPMW! Your shipping is FAST! :)
lol
glad to hear...depends on location, however... :)
dave
dave any word on more CCBBE's?