AMD presents "The Bulldozer Blog"

Printable View

Show 100 post(s) from this thread on one page

08-03-2010, 11:40 AM
JF-AMD

Don't expect performance at hot chips - that is a discussion around architecture.

As for performance, the 50% gain is an aggregate estimation of major server workloads. Estimates tend to be conservative, there is little to be gained from being overly aggressive.
08-03-2010, 11:45 AM
informal

Quote:

Originally Posted by MrMojoZ

Which has nothing to do with the issue now does it? He answered the question as much as he is going to, the repeated asking from the same poster is trying to accomplish what exactly? I know you are pro-Intel, but this should annoy anyone.

You know ,this sort of repeated asking of the same question kinda reminds me of that Simpsons episode where Marge and Homer are taking the kids to Disneyland(IIRC) and all the way Bart and LIsa keep on asking them :"Are we there yet" :D
08-03-2010, 11:50 AM
MrMojoZ

Quote:

Originally Posted by terrace215

Why doesn't repeated vacuous marketing annoy you?

Because only idiots get hyped up over pre-release marketing. I'll wait for hard data. Jeff has given us some good info at times, but when he can't give hard facts spamming him for them isn't going to help. Trying to guess the performance of up-coming products is fun but you should probably stop taking it so seriously.
08-03-2010, 11:52 AM
terrace215

Quote:

Originally Posted by Manicdan

someone asked a question and he responded with the number he is allowed to mention. i dont expect him to break NDA, do you?

My point is simply that the "number", as you say, is virtually meaningless without knowing to what it refers, so there is no point to it in isolation. I thought the example of "peak FLOPs" vs "enterprise benchmarks" was a nice illustration of the issue, as well as the "top speed" vs "mileage" car analogy.

Yet the AMD fans get so angry at a perfectly reasonable question.
08-03-2010, 11:54 AM
terrace215

Quote:

Originally Posted by MrMojoZ

Because only idiots get hyped up over pre-release marketing. I'll wait for hard data. Jeff has given us some good info at times, but when he can't give hard facts spamming him for them isn't going to help. Trying to guess the performance of up-coming products is fun but you should probably stop taking it so seriously.

Oh look! He actually answered my question.

Now, will any of the posters that were so annoyed with my question take back their complaints? :shakes:

p.s. his name is John, not Jeff.
08-03-2010, 11:54 AM
Sn0wm@n

http://www.youtube.com/watch?v=ilzuHdlKHPw LOL
08-03-2010, 11:54 AM
terrace215

Quote:

Originally Posted by informal

You know ,this sort of repeated asking of the same question

... actually got results, finally.
:shakes:
08-03-2010, 11:55 AM
Manicdan

Quote:

Originally Posted by terrace215

My point is simply that the "number", as you say, is virtually meaningless without knowing to what it refers, so there is no point to it in isolation. I thought the example of "peak FLOPs" vs "enterprise benchmarks" was a nice illustration of the issue, as well as the "top speed" vs "mileage" car analogy.

Yet the AMD fans get so angry at a perfectly reasonable question.

yes or no,
are you asking him to break NDA?
08-03-2010, 11:55 AM
terrace215

Quote:

Originally Posted by JF-AMD

As for performance, the 50% gain is an aggregate estimation of major server workloads. Estimates tend to be conservative, there is little to be gained from being overly aggressive.

Thank you, that makes the claim have meaning. Although, do you consider HPC-related work to be a "major server workload" ? I ask because that will have the biggest potential increase (~100% core improvement from 128-->256 wide), and so it obviously impacts the mean to the extent it enters into the aggregate.
08-03-2010, 11:57 AM
terrace215

Quote:

Originally Posted by Manicdan

yes or no,
are you asking him to break NDA?

He's the Director of Product Marketing for AMD servers, he can disclose what he wants to, subject to his bosses' approval, as *he* *just* *did*.
:ROTF::rofl::yepp:
08-03-2010, 12:01 PM
informal

Quote:

Originally Posted by Sn0wm@n

http://www.youtube.com/watch?v=ilzuHdlKHPw LOL

That's the one ! :)

Quote:

Originally Posted by terrace215

... actually got results, finally.
:shakes:

You really think the number he gave was because you pestered him around at few different forums(yeah SA and Ars off the top of my head).
They are starting to officially promote the new gen. of Opteron cores and it's normal they give some hints to performance uplift ,even though it is vague at this point(not surprisingly since the product is launching in 2011,duh;you could expect it was an average figure and a conservative one,just like it was the case with MC).
08-03-2010, 12:06 PM
terrace215

Quote:

Originally Posted by informal

You really think the number he gave was because you pestered him around at few different forums(yeah SA and Ars off the top of my head).

I don't post at ars, and what I think is that the fans who jumped all over me for asking the question in this thread should reconsider their attitude considering the fact that my question was answered.
08-03-2010, 12:16 PM
saint-francis

This ignore list feature is coming in more and more handy every day!

Any way, I'm still running a Q6600 because I have found no overwhelming need to move on. All I really need processing power for is X264. If I can currently encode a movie in 18 hours and a similarly clocked more modern chip can do it in 12, what does it matter? It's still going to take a whole day.
I hope that BD brings some real serious processing power to the market.
08-03-2010, 12:38 PM
JF-AMD

Quote:

Originally Posted by terrace215

Thank you, that makes the claim have meaning. Although, do you consider HPC-related work to be a "major server workload" ? I ask because that will have the biggest potential increase (~100% core improvement from 128-->256 wide), and so it obviously impacts the mean to the extent it enters into the aggregate.

HPC is a really broad category that covers hundreds of different workloads.
08-03-2010, 12:40 PM
JF-AMD

Quote:

Originally Posted by terrace215

He's the Director of Product Marketing for AMD servers, he can disclose what he wants to, subject to his bosses' approval, as *he* *just* *did*.
:ROTF::rofl::yepp:

I didn't break any NDA. I don't have an NDA, I have a confidentiality agreement.

I answered you with information that you could have found by googling, it is public.
08-03-2010, 12:57 PM
LesGrossman

Someone on this thread is having a nervous breakdown.
08-03-2010, 03:55 PM
Xoulz

Quote:

Originally Posted by terrace215

My point is simply that the "number", as you say, is virtually meaningless without knowing to what it refers, so there is no point to it in isolation. I thought the example of "peak FLOPs" vs "enterprise benchmarks" was a nice illustration of the issue, as well as the "top speed" vs "mileage" car analogy.

Yet the AMD fans get so angry at a perfectly reasonable question.

Dude, stress out, or read more and look at the architecture more and educate yourself on what those new changes might mean. Many of us have a pretty good clue..

That is why we are discussing it.

AMD is tight liped, you seem angry & want the hard facts right now...

Like a spoiled kid... = Cartman :rolleyes:
08-03-2010, 06:06 PM
freeloader

This is a rant. Not aimed at anyone who works at AMD, but at AMD's policy of not releasing any performance data a few months from a launch.

Why can't AMD throw it's customers a bone? How many people are not going to wait for Bulldozer numbers knowing that Intel is getting ready to release a new micro architecture? You're losing customers over you're refusal to even release some meaningless numbers. Intel's next architecture is set in stone now, they can't change it. About all they could do at this time is release higher clock speed parts on their 22nm process. AMD is in the same situation, although they may be able to make minor tweaks to their silicon between now and the official launch. What could AMD possibly lose right now by releasing some Cinebench or some other benchmark numbers? I'm one of the few on this board who doesn't care about architecture or the design process. I just care about performance. If you have something that can compete with Intel or even better, beat them, then let everyone know. I'd be shouting it from the rooftops to let people know about it. :shakes:

If AMD truly has a spade up it's sleeve, it's shooting itself in the foot by not playing it right now.
08-03-2010, 06:08 PM
Manicdan

in a fast moving industry like this, secrecy is very important. with a car you see prototypes cause it takes a decade for them to do anything and 9/10 things were just showing off a toy that looks just like....a toy.
08-03-2010, 06:14 PM
freeloader

Quote:

Originally Posted by Manicdan

in a fast moving industry like this, secrecy is very important. with a car you see prototypes cause it takes a decade for them to do anything and 9/10 things were just showing off a toy that looks just like....a toy.

Secrecy is very important in the initial design and testing process. When you're only a few months away from a product launch and you've got the next "big thing", let people know about it.
08-03-2010, 06:33 PM
Sn0wm@n

Quote:

Originally Posted by freeloader

Secrecy is very important in the initial design and testing process. When you're only a few months away from a product launch and you've got the next "big thing", let people know about it.

cant you just wait a couple more months before having those numbers????
08-03-2010, 06:36 PM
richierich

Yes it's only August '10, let's hope BD is released for regular desktop PCs by March/April '11.
08-03-2010, 06:40 PM
slaveondope

Quote:

Originally Posted by Sn0wm@n

cant you just wait a couple more months before having those numbers????

NOOOO:rofl: What has it been 5-7 years in the making?
I know Im stoked:yepp: #########'s umm umm um:ROTF:
08-03-2010, 07:12 PM
god_43

Quote:

Originally Posted by freeloader

Secrecy is very important in the initial design and testing process. When you're only a few months away from a product launch and you've got the next "big thing", let people know about it.

nature of the beast man, you wouldn't want amd making an announcement of an announcement...of an announcement would you p?

better to stay quiet, they are still most likely at least 5 months away....thats plenty of time for intel to muck around.

i love ignore lists!
08-03-2010, 07:13 PM
JF-AMD

Quote:

Originally Posted by freeloader

This is a rant. Not aimed at anyone who works at AMD, but at AMD's policy of not releasing any performance data a few months from a launch.

Why can't AMD throw it's customers a bone? How many people are not going to wait for Bulldozer numbers knowing that Intel is getting ready to release a new micro architecture? You're losing customers over you're refusal to even release some meaningless numbers. Intel's next architecture is set in stone now, they can't change it. About all they could do at this time is release higher clock speed parts on their 22nm process. AMD is in the same situation, although they may be able to make minor tweaks to their silicon between now and the official launch. What could AMD possibly lose right now by releasing some Cinebench or some other benchmark numbers? I'm one of the few on this board who doesn't care about architecture or the design process. I just care about performance. If you have something that can compete with Intel or even better, beat them, then let everyone know. I'd be shouting it from the rooftops to let people know about it. :shakes:

If AMD truly has a spade up it's sleeve, it's shooting itself in the foot by not playing it right now.

One word: Osborne.

Show 100 post(s) from this thread on one page