Dual Core Opteron Thread #2

Anyone know of any good tutorials on delidding dual core opterons other than the one on dfi street?

http://www.dfi-street.com/forum/showthread.php?p=539662

also, for anyone interested, the second of those two ccbbe's i had would only do 2.65 stockV.
testing at 317x9 1.465 (1.446 died after 2h40m first run, then 2h48min second run, same core).
can push it more, but will wait for a better psu first. think this will be the sweet spot though.
crapped out at 2952@1.5. would boot, but both cores would crap out after 2-10 mins one within seconds of the other...

how do these usually scale? .1v per 50mhz?

I don't know if most of you already know or not but I read something yesterday that was pretty interesting. I'll try to find the link again. Apparently the opteron isn't going to be dropping in price as much as we think. Supposedly AMD is going to make opteron 165's with only 512 cache per core instead of the 1mb cache per core. Those opterons are the ones that will be selling for about half the price. As for the 1mb cache opterons I guess they'll still drop in price but not as much as we thought. As for performance its hard to tell how much performance u'll lose with only half the cache. I guess we'll have to look at similarly clocked X2's to know.

my opty 165 ccbbe 0617FPWM at 3006Mhz :)
http://weboperative.com/gallery/Misc...113percent.PNG

CPU-Z
http://valid.x86-secret.com/show_oc.php?id=107567

had prime fail at 13h43m today(317x9@1.350x109.9%)... anyone heard of that happening before?

think it might be my underlying psu stability issue

I've had prime fail after 19hrs. I reckon it was a temp issue, just 1c too much can error out a cpu imo. If it passes 12hrs I would class that as stable tbh.

/me agrees with Ic3man
all you need is 3 rounds of prime to be stable in my book :), that's about 9-10hours

once you approach that fine line between stability and instability, prime really tells you nothing. but we all know how i feel about prime stability... so that's all i'll say :P

hey itz, just wanted to see if u flashed ur BIOS yet

well, i'm running v.3.0a

Quote:

---

Originally Posted by itznfb

once you approach that fine line between stability and instability, prime really tells you nothing. but we all know how i feel about prime stability... so that's all i'll say :P

---

->XS XTREME STABILITY TEAM
"If it ain't primestable, it ain't Xtreme"
MainRig - 523 hours, 46 minutes 13 seconds primestable<-??? :D
i was 6 days in and got an error... but i have psu issues :P looking forward to my game extreme...

also, on the topic of bioses, do they make much difference? non-manafacturer releases worth using? dangerous?

yea, that info in my sig is obviously a joke. since so many people gave us non-prime believers such a hard time,,,, absolute0 and i started the fake Xtreme stability team. obviously poking fun at the emphasis placed on prime stability

Quote:

---

Originally Posted by itznfb

yea, that info in my sig is obviously a joke. since so many people gave us non-prime believers such a hard time,,,, absolute0 and i started the fake Xtreme stability team. obviously poking fun at the emphasis placed on prime stability

---

just curious, but how do you consider a system to be stable then? im kind of starting to doubt prime/sp2004 too :stick:

this is just my opinion.... so don't start.

i prime for 2 hrs. then i do a bunch of multi-tasking tests like encrypting a couple movies while archiving/or gaming or something along those lines. i'm honestly not concerned if my PC throws errors or crashes after 10 hours of doing anything because i don't use it for that long. i game for a MAX of 2hrs at a time. but basing your stability on prime is ridiculous. you can be 48 hour prime stable, and crash in-game after 10 minutes, where temps are half what they are in prime. so what do you accomplish running prime for so long? nothing in my opinion. i think prime is a good starting off point. which i why i run for 2 hours. some times if i'm pushing my system pretty far i'll prime for 30 min with superpi running too.

blah blah blah.....

again, this is my opinion, many people agree with me, many people disagree.

Quote:

---

Originally Posted by likewhoa

heh, I hear you, but hurry damnit :p:
can you smell 3Ghz?

---

Yup, got her stable at 334x9 1.52v...w00t

man now i want a ccbbe 0617

but where to get one......

Quote:

---

Originally Posted by ineedaname

man now i want a ccbbe 0617

but where to get one......

---

I remeber newegg having them recently. I just got my ccbbe 0160 from monarch last week. Maybe you shoud try one of them.

Quote:

---

Originally Posted by itznfb

this is just my opinion.... so don't start.

i prime for 2 hrs. then i do a bunch of multi-tasking tests like encrypting a couple movies while archiving/or gaming or something along those lines. i'm honestly not concerned if my PC throws errors or crashes after 10 hours of doing anything because i don't use it for that long. i game for a MAX of 2hrs at a time. but basing your stability on prime is ridiculous. you can be 48 hour prime stable, and crash in-game after 10 minutes, where temps are half what they are in prime. so what do you accomplish running prime for so long? nothing in my opinion. i think prime is a good starting off point. which i why i run for 2 hours. some times if i'm pushing my system pretty far i'll prime for 30 min with superpi running too.

blah blah blah.....

again, this is my opinion, many people agree with me, many people disagree.

---

One of the reasons I want to be "prime stable" is because I use my PCs for distributed computing. I do not want to send back bad data, so I try and get my PC as stable as possible through a combination of SPi, Prime95, and I also run other graphics intense stuff at the same time to get the temps up and make sure I don't overheat and lockup.

Not disagreeing with you, but just stating why I think some people want their PCs "prime stable."

Quote:

---

Originally Posted by Got Chow

One of the reasons I want to be "prime stable" is because I use my PCs for distributed computing. I do not want to send back bad data, so I try and get my PC as stable as possible through a combination of SPi, Prime95, and I also run other graphics intense stuff at the same time to get the temps up and make sure I don't overheat and lockup.

Not disagreeing with you, but just stating why I think some people want their PCs "prime stable."

---

you make a good point which i'd like to add:

no matter how far you OC, you're processor is potentially going to make incorrect calculations. it doesn't matter how much you stability test there is no way to really test to see if your processor is going to make calculations as accurate as it would if it was running at stock speed. which is why you should never overclock a computer running mission critical applications. again prime and spi are going to do you no good it determining whether its going to make correct calculations all the time. that's why they sell faster processors ;)

I'm running the BOINC client with my DC project (World Community Grid), and each work unit is sent out to 3 different computers, and the results validate against each other. If the units returned are not identical, then another one is sent out to confirm. This way, the chances of using an incorrect work unit is minimized. So far on my main rig, I do not have any results which resulted in an error. Incidentally, this dual core Intel I have at work has returned several errors. It's a Dell with no OCing options and even has registered RAM!

What makes you think IT failed and not all the others? :DDD

Quote:

---

Originally Posted by Got Chow

Incidentally, this dual core Intel I have at work has returned several errors. It's a Dell with no OCing options and even has registered RAM!

---

this is my ccbwe, Yay for ccbwe's! :P

it's been running very cool and coping well with everything, no volt fluct's, nothing weird going on at all, which is good, so i decided it was time for 2.8... upped the voltage a bit, and upped the fsb and it primed for 4 mins then the readings on smartguardian started going haywire... spiraticly giving really odd readings... got a quick screeny and put it back to the time tested settings...

anyone seen anything like this before? i have had this mb to much higher fsb without any issues... though without running a divider, would that make a difference? (166)

or ldt, analog or nb voltage perhaps? all are at stock (1.2, 1.2, 1.5 in bios, though cant remember which order they sit in...)

weirdness (EDIT: shrunk the pic too much, but if you get your telescopic specs out you can jujst make out 5.75v as the reading for the 12v rail.. the ldt voltage also read 3.somthingv for a second and the temp read 159degC for a second...)

Anyone looking for an opty 165 stepping 0610DPMW? I have one on sale on ebay right now, $375 or best offer. Its a brand new OEM chip. Ive only ran it for a brief time, i tested it for DOA and stability, it ran flawless! I did a prime 95 test for 24hrs at 2.2ghz with no errors. Check out my auction if you are interested. You can pm me with questions too.

Not that this is a for sale forum but that sounds pretty expensive (to me at least) for a chip that does not guarantee a specific clock... But more power to ya, I'm sure somebody would like that stepping.

Quote:

---

Originally Posted by adamryer

Anyone looking for an opty 165 stepping 0610DPMW? I have one on sale on ebay right now, $375 or best offer. Its a brand new OEM chip. Ive only ran it for a brief time, i tested it for DOA and stability, it ran flawless! I did a prime 95 test for 24hrs at 2.2ghz with no errors. Check out my auction if you are interested. You can pm me with questions too.

---

hmm...i dont think anyone in their right mind would pay $375 for a CCBBE now. i barely sold mine for $300 and it did 2.85 GHz. if u want a better sale. test it 2.8+. noone runs their opty at 2.2

EDIT: i mean noone XTREME runs their opty @ 2.2 ;)

yea, the Opty completely lost its value when conroe became redily available, and X2's prices went through the floor.