BSN article:AMD Insiders Speak Out: BAPCo Exit is An Excuse for bulldozer performance

[Dribble]

Interesting article on BSN:
http://www.brightsideofnews.com/new...-for-poor-bulldozer-performance.aspx?pageid=0

Looks like a frustrated engineer upset at executives more interested in FUD then trying to produce a good chip.

 

[3DVagabond]

What I wan to know is how AMD convinced nVidia and VIA to endorse their FUD and quit as well.

 

[Terzo]

Does BSN have an editor? There are so many errors (grammatical) in that article that I doubt anyone bothered to proofread it.

If true, I wouldn't be surprised about the article. It sounds like such wheeling and dealing is par for the course course for the industry.

 

[ShadowVVL]

There are so many errors (grammatical) in that article that I doubt anyone bothered to proofread it

maybe he was in a rush to get it posted.

its alot of reading and its 5 am so ill finish it tomorrow. thank you for posting dribble, it might be the most interesting thing ive had to read this year.

 

[dma0991]

I'm a little skeptical about BSN and Theo Valich after the supposed 'Bulldozer not delayed drama'. There may be some truths to the article but I'd still consider it a FUD since AMD was not the only one walked out of BAPCo.

Zacate did not disappoint, so did Llano and hopefully Bulldozer will not disappoint as well. For a site that names itself bright side of news, the news is kinda dark and gloomy.

 

[Ancalagon44]

Doesnt explain why Nvidia left at exactly the same time.

 

[notty22]

There was probably a common deadline for commitment/payment.
Its not the least bit surprising to read claims that AMD has its own stealth anti-marketing, that some believe only Intel/Nvida uses.

When asked about core performance, surprising information was that a Bulldozer core versus the existing cores in Llano will result in minimal improvements overall. Our sources went on to say that the launch of Llano clearly shows what is the current and future strategy - downplay CPU performance every chance you get. Everything has to revolve around the GPU.

 

[yottabit]

Oh God, I thought it was an interesting read until you guys pointed out it was Theo Valich. He's the same one that claimed that AM2 would have "Reverse Hyper-Threading"
http://www.theinquirer.net/inquirer/news/1005560/amd-socket-am2-has-a-secret-weapon

I would treat anything he'd write much like you'd treat a tabloid

 

[daveybrat]

Oh God, I thought it was an interesting read until you guys pointed out it was Theo Valich. He's the same one that claimed that AM2 would have "Reverse Hyper-Threading"
http://www.theinquirer.net/inquirer/news/1005560/amd-socket-am2-has-a-secret-weapon

I would treat anything he'd write much like you'd treat a tabloid

Lol, wow, that article on Reverse Hyper-threading is a riot. Ridiculous!

 

[Stuka87]

Majority of it is an interesting article. I just don't know how much of it to believe.

 

[Axon]

I guess we'll see. A lot of what AMD's behavior seemingly confirms this story, but we can't really tell until the benchmarks.

 

[Idontcare]

The "story" is obviously plausible (it could happen)...but it is up to each of us to decide for ourselves just how probable it is.

I have my doubts if for no other reason than the fact that management always intentionally withholds information from the engineering groups.

Theo may well have tapped into some legitimately soured AMD engineers but that does not mean those engineers are actually privy to the real-deal that is going on in discussions at the executive level.

What sets off my "red flag alert" regarding the "sources" is that the majority of the content of their quotes, as well as the majority of the content of Theo's added article text, is just your usual garden-variety hyperbole and rhetoric.

For example:

"Bulldozer is going to disappoint people because we did not get the resources to build a great CPU, and it's not that we needed billions of dollars to make it a leader. We needed investment in people, tools and technology."

...well let's be honest, this could be said about anything.

What engineering project exists that could not stand to benefit from added resources?

I'm pretty confident that Intel engineers could make a convincing argument that Sandy Bridge could have been an even better CPU if only Intel invested more in people, tools and technology.

And let's be pragmatic about this, its one thing to say "if only we had more things that cost money" and quite another to have the money to spend on securing those resources.

Look at AMD's balance sheet, not sure where these particular engineers think AMD would have come up with the dough...you can't squeeze blood from a stone.

The gist of the story might be true, its plausible, if you can't beat the competition then you do need to do something...but the basis for the story sounds, to me anyway, like a handful of unrealistic (and possibly ill-informed) engineers who want to blame today's management for not changing reality by waving around some magical resource wand.

 

[yottabit]

It's classic sensationalist journalism to mix in real (or at least plausible) information with complete utter bull and wild guesses. That way if some of the wild claims end up being completely untrue, the author can point back that they were right about some of the points in his article for damage control

The other thing that bothers me about the article is this engineer (who is such a loyal employee and will stay with the company forever) is breaking all sorts of rules by doing this kind of leak, and he wants to stay with the company forever? It doesn't make sense, and it's not like he would be helping the company by doing this- he would only ever be hurting it.

 

[lifeblood]

I'm sure AMD marketing does try to downplay any information that shows AMD in a bad light. Its part of their job. I also sure the Engineers were not getting everything they wanted/needed to make good CPU's, AMD was deeply in debt until recently because of its purchase of ATI and didn't have a lot of cash. AMD, like any company, has certainly been playing down their weaknesses and promoting their strengths, all the while working on a product that will get them on top again. AMD has been using innovation in the form of APU's to counter balance other problems.

I'm sure the engineer in telling the truth as he sees it from his perspective. That doesn't make it completely factual, just true as he sees it.

 

[exar333]

It sounds like frustration from within a company that has spent a lot of time and money on something that may not live to up expectations. Whether this is true or not, time will tell. If so, this is very disturbing news from AMD.

I am sure the 'zoners will spin this as a 'planted Intel spy' undermining AMD morale and such.

 

[bridito]

My humble opinion: I am committed to believing only extensive benchmarking performed by reputable and unbiased third parties on any CPU, whether AMD, Intel, ARM, or Jersey Shore Silicon Inc. Everything else is just white noise.

 

[KingstonU]

LOL at Jersey Shore Silicon Inc.

But seriously are we next going to hear next that bulldozer is going to slip into 2012? "The waiting is the hardest part...."

 

[formulav8]

blah blah blah nvidia and via left because of bd as well. I personally just don't much care either way when I think about it

 

[Cerb]

So why not ditch Sysmark years ago, then? Can you not but laugh at a C2Q 2.4 keeping up with a AII 3Ghz in Sysmark, but practically nothing else? Even if BD flops, it doesn't make sense, as Sysmark favoring Intel by more than their actual performance advantage is anything but new.

 

[DaveSimmons]

Via and nvidia were disappointed by Bulldozer performance? Who knew!

It might be plausible, but then we can also make up plausible stories about intel paying BAPco under the table to skew benchmarks in their favor. It could happen, so maybe it did.

 

[Idontcare]

So why not ditch Sysmark years ago, then? Can you not but laugh at a C2Q 2.4 keeping up with a AII 3Ghz in Sysmark, but practically nothing else? Even if BD flops, it doesn't make sense, as Sysmark favoring Intel by more than their actual performance advantage is anything but new.

I can think of a few reasons why this would be the case and AMD would still "stick around" for a while.

The recent'ness of the decision to leave appears to have been triggered by the milestone of Bapco having finalized and "gone gold" the Sysmark2012 bench suite. Once that threshold was crossed there was no plausible way for AMD (or NV or Via) to continue internal lobbying and negotiations to get Sysmark2012 weightings changed in their favor.

As for why now versus 4 yrs ago...I suspect that does have something to do with llano and bulldozer. They held out to the bitter end with the internal negotiations to get Sysmark setup to be as favorable to their terms as possible. This is obvious good timing for AMD.

Why Via and NV both decided to pull out now, versus 4 yrs ago when the scales will still equally tipped against their favor, no idea. Maybe it was intentionally coordinated by AMD to maximize the PR effect and mindshare effect?

Via and NV had nothing to lose by leaving Bapco and nothing to gain by sticking with Bapco.

If NV and Via left (which they did) then they at least stood/stand a chance of gaining something in terms of positive PR. The status quo obviously wasn't getting them anywhere, AMD is letting them know in advance they are going to leave Bapco and make waves in the media about it, so why not jump on the bandwagon and see what shakes out?

There is another possibility, we are all speculating that AMD left in advance of rolling out bulldozer because bulldozer will be lacking in this bench, but what if they left because Intel is getting ready to roll-out IB (or MIC, or Atom on 22nm) and they just absolutely slaughter the competition with these new designs. That would go some distance to explaining why NV and Via suddenly have an interest in jumping ship.

It's all conjecture, only Bapco knows the truth and I doubt Intel is going to let them open that kimono

 

[veri745]

I feel like I could have written this article, purely based on things that were "probably" happening with AMD falling behind Core2 and then was subsesquently broke after the ATI acquisition.

Throw in an agenda to discredit Bulldozer, use AMD's recent resignation of BAPCo, and you end up with a totally generic but plausible story about how Bulldozer might suck.

The only really "new" piece of info I learned is that AMD was pushing SYSmark while they had superior products but were being shut out of the market by Intel tactics. This is totally unsurprising, though.

 

[Arkadrel]

It's classic sensationalist journalism to mix in real (or at least plausible) information with complete utter bull and wild guesses. That way if some of the wild claims end up being completely untrue, the author can point back that they were right about some of the points in his article for damage control

The other thing that bothers me about the article is this engineer (who is such a loyal employee and will stay with the company forever) is breaking all sorts of rules by doing this kind of leak, and he wants to stay with the company forever? It doesn't make sense, and it's not like he would be helping the company by doing this- he would only ever be hurting it.

Agree with above, the artical sounds like it was written for the site hits it would generate.

I seriously dont think they have anyone quoted for that, which means = artical is made up, quotes are made up.

Stuff like that kills your reputation.

 

[GaiaHunter]

16 of July there will be the [H] lan (http://hardforum.com/showthread.php?p=1037335089#post1037335089) so maybe we will hear more about BD then and I wouldn't show a dog with Kyle around especially if the release is only 2 months later or 2012.

 

[Cerb]

I can think of a few reasons why this would be the case and AMD would still "stick around" for a while.

It's all conjecture, only Bapco knows the truth and I doubt Intel is going to let them open that kimono

All but the last one make sense. That IB will beat BD [on the desktop, if you are willing to be spendy, ]is all but guaranteed (next gen Atom would have been figured out before Brazos' performance was known, I would think, and Intel may not really care to go against Brazos, so I'll leave that up the air). That would not be at all unfair to AMD, nV, nor VIA. Sysmark 2012 ending up as another iteration of "Intelmark" OTOH, does make sense, as Llano, FI, actually fares far better in real applications than Sysmark scores would have anyone believe; and, AMD, VIA, and nVidia would all have much to gain by GPU-acceleration going into the scores, even with plain old DirectDraw.