Dual Core Opteron Thread #2

[Russ_64]

I heard that AMD has stopped production on all S939 CPUs ?

[Classic Satch]

I heard that AMD has stopped production on all S939 CPUs ?

Yep, S939 is officially retired. So if any of you need some parts, I would suggest grabbing them now.

[QuietIce]

I think 4 CPUs is enough to play with for a little while ...

[Kimball]

Hey I was just reading around here again, and noticed you talked about some TCCD syndrome... Can you explain that one for me

the short: they wont do cas 2.5 anymore, which sucks since they did 320mhz @2.5-4-4-10 one year ago :/ Hit a search at tccd syndrome and ull find some threads about it.

Updates:
SuperPi 1M >27s


Highest bootable: 3276Mhz
http://valid.x86-secret.com/show_oc.php?id=171350

[Rev.]

I picked up this:


I am air cooling with a Tuniq Tower and have not taken the IHS off.

I run it at 250 x 12 at default voltage.

[turbox997]

Yep, S939 is officially retired. So if any of you need some parts, I would suggest grabbing them now.

Where's the proof? I've seen EOL charts for s939 A64 X2's, but not dc Opterons, single cores yes...

This is a server proc, and any business that actually depends on this would expect it to have some stock left over for a while...without having to upgrade a whole new platform.

[bt_medic04]

I picked up this:
[Monstrous Picture]

I am air cooling with a Tuniq Tower and have not taken the IHS off.

I run it at 250 x 12 at default voltage.

nice stepping

[Rev.]

Sorry about tha monstrous picture. I have considered removing my IHS, but I am not sure what I would gain. 3.0Ghz is a wall of sorts for me. Adding voltage doesnt seem to help get a higher clock (at least on air). Would I see a good chance or more clock speed without the IHS?

Also, does anyone know where CPU "dot" pads (like the ones on the XPs) that would help protect the core from the pressure of this Tuniq Tower if I did take the IHS off?

[nick2crete]

Sorry about tha monstrous picture. I have considered removing my IHS, but I am not sure what I would gain. 3.0Ghz is a wall of sorts for me. Adding voltage doesnt seem to help get a higher clock (at least on air). Would I see a good chance or more clock speed without the IHS?

Also, does anyone know where CPU "dot" pads (like the ones on the XPs) that would help protect the core from the pressure of this Tuniq Tower if I did take the IHS off?

Dont do it..doesnt worth the trouble to get only 20MHz more ,void your warranty..etc..
You have a nice cpu with a very good o/c result ..

[biohead]

Sorry about tha monstrous picture. I have considered removing my IHS, but I am not sure what I would gain. 3.0Ghz is a wall of sorts for me. Adding voltage doesnt seem to help get a higher clock (at least on air). Would I see a good chance or more clock speed without the IHS?

Also, does anyone know where CPU "dot" pads (like the ones on the XPs) that would help protect the core from the pressure of this Tuniq Tower if I did take the IHS off?

What are your temps at load? And what are your MAL (Max Async Latency) and RP (Read Preamble) set to?

[Rev.]

What are your temps at load? And what are your MAL (Max Async Latency) and RP (Read Preamble) set to?

My Max load is around 50C, but my temp readings have been skewed since I put the 185 in, so I dont believe them. The other 2 settings I will check when I get back later in BIOS.

http://i2.photobucket.com/albums/y36...ator/3ghzp.jpg

I primed last night for 10+ hours and here is the screenie.

EDIT #1 My MAL and RP are both set to auto in BIOS, I have no clue where to even begin with those settings.

EDIT #2 I dowloaded A64 tweaker, my MAL is 7ns and my RP is 5.5ns

Would loostening these give me more bandwidth possibilities?

[Lastviking]

My Max load is around 50C, but my temp readings have been skewed since I put the 185 in, so I dont believe them. The other 2 settings I will check when I get back later in BIOS.

http://i2.photobucket.com/albums/y36...ator/3ghzp.jpg

I primed last night for 10+ hours and here is the screenie.

EDIT #1 My MAL and RP are both set to auto in BIOS, I have no clue where to even begin with those settings.

EDIT #2 I dowloaded A64 tweaker, my MAL is 7ns and my RP is 5.5ns

Would loostening these give me more bandwidth possibilities?

change to MAL 9 and RP 6.5 that can fix your max wall.

Damn i want your cpu...or one with same stepping...

[GUFY]

My CPU have cold-bug at 250 FSB and cpu temp -5.
I hate cold-bug at AMD cpus

[MadDias]

hi guys,
i´d like to see the stepping of my opty165 on that list in the first post of this thread. Almost everyone i´ve seen with this stepping is hitting 3ghz.
it´s a ccbbe 0616xpmw
334x9@1.54v (24/7 setup)

here´s a link to my ocdb-entry on diy-street for a 8hrs Orthos screenshot and some 3d benchmarks:
http://www.diy-street.com/forum/show...&postcount=379

[Final8ty]

Sorry about tha monstrous picture. I have considered removing my IHS, but I am not sure what I would gain. 3.0Ghz is a wall of sorts for me. Adding voltage doesnt seem to help get a higher clock (at least on air). Would I see a good chance or more clock speed without the IHS?

Also, does anyone know where CPU "dot" pads (like the ones on the XPs) that would help protect the core from the pressure of this Tuniq Tower if I did take the IHS off?

I have the same stepping CCBBE 0613RPMW

IHS ON


IHS OFF and notice the 8c & 10c drop on both cores from the first pic.

Been upto 3120 at 1.360v

[biohead]

whaaaaaaaaaaaaaaaa?

that chip is AWESOME!

what are your idle temps?

[Rev.]

Yeah I just cant keep mine prime Stable on air at 3120mhz.

I fear the mass and weight of the tower I have cooling will crush the core if I removed the IHS.

BTW mine is is 39C right now....that is with AVG doing a virus scan.


Last night I read how Orthos prime is supposed to be hard...hard enough to fail normal prime stable setups....So I had to take the challange at bone stock. BTW I cant go any lower on vcore per my motherboard....so I have no clue if it will do this at 1.26v.....but I suspect it would.
Large Pic Warning:
http://i2.photobucket.com/albums/y36...rthospwned.jpg

[Classic Satch]

Where's the proof? I've seen EOL charts for s939 A64 X2's, but not dc Opterons, single cores yes...

This is a server proc, and any business that actually depends on this would expect it to have some stock left over for a while...without having to upgrade a whole new platform.

I was referring to an article in Maximum PC magazine. They may have been referring only to the "desktop" chips like the X2's and FX series. Sorry for the confusion.

[Classic Satch]

Yeah I just cant keep mine prime Stable on air at 3120mhz.

I fear the mass and weight of the tower I have cooling will crush the core if I removed the IHS.

BTW mine is is 39C right now....that is with AVG doing a virus scan.


Last night I read how Orthos prime is supposed to be hard...hard enough to fail normal prime stable setups....So I had to take the challange at bone stock. BTW I cant go any lower on vcore per my motherboard....so I have no clue if it will do this at 1.26v.....but I suspect it would.
Large Pic Warning:
http://i2.photobucket.com/albums/y36...rthospwned.jpg

I would suggest running Orthos at Priority 10. It puts one hell of a strain on your system. It will show errors much more quickly than running it at Priority 1. I've had some configurations Orthos stable for 10 hours+ with Priority 1 but when I learned that the priority numbers are actually inverse from what I assumed, the same setup would fail in under an hour running Priority 10.

[Rev.]

I would suggest running Orthos at Priority 10. It puts one hell of a strain on your system. It will show errors much more quickly than running it at Priority 1. I've had some configurations Orthos stable for 10 hours+ with Priority 1 but when I learned that the priority numbers are actually inverse from what I assumed, the same setup would fail in under an hour running Priority 10.

Orthos priority 10 is insaine...there is no real life application that I would ever use that would do this to my system!!!!

That being said...30 mins and no errors so far

[halo112358]

Hi, just a note to the "I'm afraid of chipping my die" folks - unless you're hauling your box to lan parties and literally slinging the machine into the back seat of your car you're not likely to crush a die. Those copper shims they sold us for older athlons, they weren't neccessary either ^^

[Chain]

First post here so go easy on me.

I recieved an Opty 165 LCB9E 0636 on the 15th of this month from Newegg's CA warehouse to replace my single core venice 3000+ (Damn you Supreme Commander). I've been doing a lot of research and haven't found any real successful overclocking stories with this stepping, and it's not looking good for me breaking that pattern.

I decided to pop off the IHS to begin with in an attempt to get the lowest possible temps I could right from the start. It easily cruised up to 2.7ghz (300x9) at default voltage (1.35v). Passed that I really haven't had much luck at all. I have it running Orthos right now at 2.745 at 1.4V but I have extreme difficulty even loading Windows beyond 305 HTT. Temps under load at 2.745ghz are 42C/31C for the individual cores and 35C mb sensor using Everest (it reads exactly the same as coretemp). The individual core temps seem to be bugged to, since at idle core1 sits well below room temperature.

I've kept the RAM running at or below 250mhz using mostly the 5/6 and 3/4 dividers. I've tried up to 1.55V just to load into Windows at 310 HTT but its mostly been a no go. I've only had one successful windows boot at 310 HTT with the command rate set to 2T, but it crashed within a minute of running Orthos. I installed all the drivers/hotfixes/optimizers and it didn't make any difference in my stability.

Also, I tried dropping the multiplier to 8x and running above 305 HTT but I still got the same results. So I'm thinking that there may be something else holding this chip back (atleast I hope) that I am not seeing. But, I didn't have this same problem with my 3000+, I could boot into Windows up to 315 HTT. I'm running the latest BIOS from DFI, I think its from 6/06.

So if anyone has some suggestions on how I can break this wall down I'd appreciate it. The chip wants to do 2.8+, I just know it. I can post some more info/screens whatever when I get home (damn 12 hour shifts). Thanks!

[RazorBoy]

My 165 opty shots cooled with Air:
8M x 2 @ 3004MHz
2664MHz @ 1.26v

2700MHz @ 1.26v =)

[Rev.]

I decided to pop off the IHS to begin with in an attempt to get the lowest possible temps I could right from the start. It easily cruised up to 2.7ghz (300x9) at default voltage (1.35v). Passed that I really haven't had much luck at all. I have it running Orthos right now at 2.745 at 1.4V but I have extreme difficulty even loading Windows beyond 305 HTT. Temps under load at 2.745ghz are 42C/31C for the individual cores and 35C mb sensor using Everest (it reads exactly the same as coretemp). The individual core temps seem to be bugged to, since at idle core1 sits well below room temperature.

Hey Chain, I had a bad stepping 175 that topped off at 2.6. So a 165 at 2.7 aint bad.

[Rev.]

I am using air....