Here's a little teaser....

[s7e9h3n]

i was hinting that performance was markedly improved vs. the B1 benchmarks that sheeple had been mindlessly quoting as production performance.

Dave: In AMD nomenclature, *1 steppings are no different than any other steppings in terms of performance. That's a trait that's reserved for Intel procs . In fact, some of the best performing AMD cpu's EVER have been *1. If people had been quoting b1 performance, those procs will be exactly the same clock for clock as retail

[dave_graham]

Dave: In AMD nomenclature, *1 steppings are no different than any other steppings in terms of performance. That's a trait that's reserved for Intel procs . In fact, some of the best performing AMD cpu's EVER have been *1. If people had been quoting b1 performance, those procs will be exactly the same clock for clock as retail

http://www.xtremesystems.org/forums/...&postcount=150

go back one post.

B1 is DVT and not production release.

dave

[s7e9h3n]

http://www.xtremesystems.org/forums/...&postcount=150

go back one post.

B1 is DVT and not production release.

dave

Yes, each successive stepping makes improvements on the cpu's but - like you say - the improvements are most likely reserved to fix erratas. Even production/retail cpu's contain erratas which may or may not be corrected. That being said, performance in b1 cpu's are basically identical to successive steppings with the exception of performance in VERY specific applications. B1 is not necessarily reserved for DVT as they're found in many OEM machines as well. By chance, do you have a pic of your cpu's? Just want to compare "notes" vs. the ones I have on hand

[dave_graham]

Yes, each successive stepping makes improvements on the cpu's but - like you say - the improvements are most likely reserved to fix erratas. Even production/retail cpu's contain erratas which may or may not be corrected. That being said, performance in b1 cpu's are basically identical to successive steppings with the exception of performance in VERY specific applications. B1 is not necessarily reserved for DVT as they're found in many OEM machines as well. By chance, do you have a pic of your cpu's? Just want to compare "notes" vs. the ones I have on hand

all my cpus are thermal'd but the 2347s are week 27 and the 2344s are wk 24

cheer,s

dave

[w0mbat]

And which Stepping from 0724/27? I´ve got a B1 w/ 0730.

[dave_graham]

And which Stepping from 0724/27? I´ve got a B1 w/ 0730.

from memory, don't know.

when i get home.

dave

[s7e9h3n]

all my cpus are thermal'd but the 2347s are week 27 and the 2344s are wk 24

cheer,s

dave

When you get a chance though, could you let me know the info on the ENTIRE second line....that info helps more than the production date. Thanks

Oh, btw, here are a couple examples of stepping *1 cpu's that I've owned and were decent performers. (I know you don't care too much for OC'n, but these are the only examples that I have )

Here's the former WR dual core - which to this day is STILL one of the strongest dual core's I've ever seen:




My "FX56" - Yes, this one WAS a DVT cpu:



Lastly, here's the only 4g Opteron.....ever (at least last time I checked) :



I've got a few more, but I really don't feel like looking through my bag of screenshots

[Lightman]

Guys do you have any idea what vCore Opteron 2360 sent to Anand has? Is it as ridiculous as on CPU-Z screen in their review? (1.520V!)

Also same CPU-Z screen is showing rev. B2 for 2.5GHz K10

A handy LINK

[s7e9h3n]

Guys do you have any idea what vCore Opteron 2360 sent to Anand has? Is it as ridiculous as on CPU-Z screen in their review? (1.520V!)

Also same CPU-Z screen is showing rev. B2 for 2.5GHz K10

A handy LINK

Not sure, but I'm getting tired of waiting for Asus....I think I'll just run this pair in a Tyan s3992....Gonna go pick up the board in a bit. I'll be able to compare tonight

[PhilDoc]

Sounds like fun, after you're done send me the cpus, lol. Enjoy, hope you get some good tests done.

[misteroadster]

Waiting for your results

[Noob-ftw]

Guys do you have any idea what vCore Opteron 2360 sent to Anand has? Is it as ridiculous as on CPU-Z screen in their review? (1.520V!)

Also same CPU-Z screen is showing rev. B2 for 2.5GHz K10

A handy LINK

1.5V for 2.6ghz is ridiculously high. I was expecting a max of 1.4V...

Hopefully it's because Anand doesn't know how to overclock.

[PhilDoc]

1.5V for 2.6ghz is ridiculously high. I was expecting a max of 1.4V...

Hopefully it's because Anand doesn't know how to overclock.

you can't go by cpu-z, it still has trouble reading my 185, let alone a new cpu.

[s7e9h3n]

http://www.xtremesystems.org/forums/...&postcount=150

go back one post.

B1 is DVT and not production release.

dave

Just confirmed that B1 IS the stepping on currently available Barcelonas. We won't be seeing B2's until AMD starts ramping up the speeds on quads; as evidenced by Anandtech's EARLY Phenom chip.

[Periander6]

Just confirmed that B1 IS the stepping on currently available Barcelonas. We won't be seeing B2's until AMD starts ramping up the speeds on quads; as evidenced by Anandtech's EARLY Phenom chip.

But that Dave Graham guy was swearing up and down that B1 had some performance crippling bug, how could they be shipping that? Or is the Dave Graham guy full of crap?

Oh wait, here was his original post on the subject:

http://www.xtremesystems.org/forums/...&postcount=150

So that performance crippling bug that he claimed invalidated the the Coolaler benchies was a bug that impacted a single set of obscure memory bandwidth benchmarks (http://www.streambench.org/) that I'd never even heard of before.

LOL

[Sparky]

you can't go by cpu-z, it still has trouble reading my 185, let alone a new cpu.

Agreed, on new hardware it tends to not be accurate. It had the hardest time with my 165 for awhile until a few updates ago, even so it still kinda bounces around some.

[dave_graham]

i'm sorry, but AMD has stated (at least in SEVERAL emails) that BA steppings are production. whether that's a kludged B1 is beyond me, i'm simply reporting what i've been told.

if that makes me "full of crap," Periander, then so be it. Blindly trusting engineers that live off this stuff day to day is not something that i do, but I do trust those who use this stuff day to day.

I didn't pull the BA stepping out of my ass. It was specifically stated that:
a.) the B1 stepping was not production (could be my week, sure, but...i digress)
b.) that BA/B2 steppings solved a significant errata present in the B0/B1 steppings (aka DVT spins) that increased the nominal performance of the Barcelona in STREAM processes

so, we'll see...won't we? as it stands, AMD is at least 3-4 weeks off from providing the channel with parts, so....since all of August isn't accounted for yet, we still might see some changes..

anyone want to gank the microcode from my BIOS's? that might answer your questions since the northbridges (or MCs) are identified by stepping (at least on the S3992)

cheers,

dave

[red]

Why would AMD give TR, AT, various other sites B1 if they don't intend to ship that? That doesn't make any sense.

Maybe when Gary Key said it flies after 2.4GHz, B1 will be under 2.4, B1+ will be 2.4+ (with added IPC boost from various things, errata fix). That's my logic

[Periander6]

b.) that BA/B2 steppings solved a significant errata present in the B0/B1 steppings (aka DVT spins) that increased the nominal performance of the Barcelona in STREAM processes

So what?? As far as I can tell, none of these reviews you claim were invalidated even ran these damn STREAM benchmarks nobody's ever heard of. STREAM is just an obscure benchmark suite that AMD cherrypicked to show off because it's a pure memory bandwidth bench, the area where AMD is strongest. What the hell difference would this errata have when running non-STREAM benchmarks? How would this justify your questioning the results IBM sent up to Spec based on an errata that has nothing to do with any of the Spec benches like you did here?

http://www.xtremesystems.org/forums/...4&postcount=41

As if IBM were a bunch of utter morons posting Spec results for a production server not shipping until mid-November but fitted with a damn DVT processor with performance crippling errata? WTF?

[tsuehpsyde]

Not sure, but I'm getting tired of waiting for Asus....I think I'll just run this pair in a Tyan s3992....Gonna go pick up the board in a bit. I'll be able to compare tonight

Best of luck on getting us some scores to look at, and hopefully some results as to how these chips overclock.

[PhilDoc]

Best of luck on getting us some scores to look at, and hopefully some results as to how these chips overclock.

Since AMD shipped some 2.5G, I suspect there will be some that will do 2.5G on stock cooler and stock voltages. Probaby not many, but some. Than'ts why Anands chip was hand written.

[ShowsOn]

I dropped in the forums here yesterday to see what the buzz was about quad AMD's....sadly, there isn't all that much I guess it's time to try to make some waves Mind you, this is only a pic of a pair of quads I've got here (with almost any identifying info photochopped out by me ). Note the pack date on the chips: ~ a little over one month ago. Trust me, these are the real deal....I'll post some screens when/if I ever get them to recognize correctly on my L1N64. They do, however, run fine on a few server-class boards but what fun is overclocking using registered ecc mem ?

Hi, if you are willing to run some benchmarks on these CPUs, I'd love to see how they go on the x264 MPEG4-AVC video encoder. There are some benchmark files at doom9.org that enable a standardised test (i.e. specific versions of x264 and the file to encode). See here:
http://forum.doom9.org/showthread.php?t=122318

I'm most interested in seeing how ONE Quad Core Opteron runs. There are some 2 x dual core opteron benchmarks at doom9.org that it can be compared against.



Thanks for your time.

[STEvil]

Hey ShowsOn, this looks like a rip-off of that link you posted..? http://www.xtremesystems.org/forums/...d.php?t=158198

[s7e9h3n]

i'm sorry, but AMD has stated (at least in SEVERAL emails) that BA steppings are production. whether that's a kludged B1 is beyond me, i'm simply reporting what i've been told.

if that makes me "full of crap," Periander, then so be it. Blindly trusting engineers that live off this stuff day to day is not something that i do, but I do trust those who use this stuff day to day.

I didn't pull the BA stepping out of my ass. It was specifically stated that:
a.) the B1 stepping was not production (could be my week, sure, but...i digress)
b.) that BA/B2 steppings solved a significant errata present in the B0/B1 steppings (aka DVT spins) that increased the nominal performance of the Barcelona in STREAM processes

so, we'll see...won't we? as it stands, AMD is at least 3-4 weeks off from providing the channel with parts, so....since all of August isn't accounted for yet, we still might see some changes..

anyone want to gank the microcode from my BIOS's? that might answer your questions since the northbridges (or MCs) are identified by stepping (at least on the S3992)

cheers,

dave

Dave, it wasn't my intention to insinuate you were lying. I just think you or maybe your FAE may be mis-informed. I feel I have to clarify a couple of things though:

1. There isn't such a thing as a "BA" stepping. Every stepping I've ever seen (along with everybody I've spoken to about this issue) is a letter followed by a number. Maybe it's not the stepping your speaking of...instead, could it be the revision #?

2. Steppings aren't identified by production date. As a matter of fact, the production date doesn't give too much info about the cpu itself. Rather, it's the first five letters/numbers found on the second line of AMD cpu's which give much more pertinent info than the date of manufacturing.

3. Almost ALL final/production level hardware - especially cpu's - contain erratas. These erratas are deemed "acceptable" by the manufacturers as they only affect VERY SPECIFIC situations. If there's a huge problem in the chips, they simply won't be released -> as with the case of the B0 Barcelonas. Successive revisions of the cpu's are designed to fix certain erratas, but clock-to-clock performance shouldn't be affected.

On another note, I wasn't able to pick up the s3992 as I had originally planned..... .....

[kyosen]

AMD officially uses the expression "Stepping BA" on its site:
http://products.amd.com/en-us/OpteronCPUFilter.aspx
but yeah I think that it may differ from CPU-Z showing.
On Anandtech, K10 2.5GHz ES shows stepping 2 on CPU-Z,
and revision is grey-out.
I suspect 2.5GHz ES may be "Stepping BA", only just speculation.

Anyway I still have a little, a very little hope for newer stepping & BIOS