The AMD Execution Thread [2007 - 2017]

[no-X]

This situation is different - really surprising. It's hard to tell anything unless we know Eric's reasons...

 

[Alexko]

So the Internet was blowing a fuse over Mr. Killebrew leaving AMD, but it seems to be quite chilled out about Eric not being there anymore? Boy, the Internet sure knows its stuff!

Well, Carrell Killebrew appeared to have been fired, while it seems Eric Demers decided to leave. Of course, we don't actually know anything for sure, so…

 

[AlexV]

This situation is different - really surprising. It's hard to tell anything unless we know Eric's reasons...

What is exactly different about it? You have people leaving a company. One is a key asset that had an active involvement in the core development of most of ATI's architectures, and one of the more brilliant engineers in the business. The Internet doesn't care because...he wasn't fired? Err...this is wrong. I bet that if there was a soap-operaish piece about how Eric solved something in a single night, with a cup of coffee and some of his buddies around a table, there'd be a far stronger reaction. I hope I am wrong.

Here's something that I'm considering when looking at this: Bergman left out of his own accord, as did Pat Gelsinger earlier, as did Huddy, as did quite a few other people in the business along the years. It seems to me that people tend to leave to "pursue other opportunities" when they either lose a big organisational gamble (Bergman, Gelsinger) or when they become disgruntled with their organisation / feel it's no longer worth their time. And the latter would be quite worrying in context. If Joe Macri leaves soon too (I think he's still there), the future prospects would seem quite...worrying, in terms of retaining relevance. But that's just MHO.

 

[Kaotik]

TBH I think there's far simplier reason why Dremers might have been "disgruntled with his organisation" - he might feel that AMD just "passed him" when recruiting Papermaster as the CEO, which could easily give motivation to jump somewhere else, and not having anything to do with "AMDs future outlook"

 

[3dilettante]

Maybe people are burnt out on the news of yet another senior executive leaving? It's a well-worn pattern by now.

Let's look on the bright side. If AMD manages to get its HSA initiative accepted by other vendors (if), and Demers remains in the field (seems likely), GCN and its ilk might not be the last time that a Demers-associated design winds up on an AMD chip.
Given the serious headwinds AMD is facing, in this hypothetical future he might be interacting with his former bosses from a more advantageous position.

AMD has some assets that, while being second to Intel and possibly redundant to three or so ARM architecture licencess, could be intriguing to a sliver of possible customers.
There are some murmurs of a possible realignment or change of the graphics group's focus and schedule, which would be a tiresome repeat of a trend that indicates that leadership or competitiveness in anything, particularly if it requires investment and not free-riding on Intel, is best to be avoided.

I had written more about AMD's headwinds, but thought the better of it.
I'm starting to suspect it might become boring to me, and I don't feel like going into depth on that.
It's not their job to be fascinating to me, obviously, but I don't post about Via or Matrox much, either.

 

[Silent_Buddha]

IMO, Eric Demers leaving is the worst blow of all the former ATI people leaving or being let go.

But I'm more interested in learning of his reasons and motivations for leaving rather than make a ruckus about it.

Was he just tired and wanted something new? Was there politics involved and an internal power struggle over which direction the GPU business would be going? Did another company offer him double his current salary? Etc...

I like Sireric. Especially the handle. Miss seeing his posts. Miss seeing him at AMD/ATI. But the world moves on. I can only hope it wasn't due to Eric thinking AMD were headed down the wrong path with regards to the GPU business unit and that AMD remains competitive there.

Regards,
SB

 

[Dave Baumann]

Read the reports from the analyst day, especially Lisa Su's comments with regards to graphics.

 

[trinibwoy]

This?

Ask Read and he'll tell you that's a key advantage for AMD, that its CPU and GPU IP will bring more value through a better overall experience in the market. That's a strategy less obsessed with raw specs and sheer speed and more focused on a holistic package. Senior VP Lisa Su said AMD will aggressively enter the tablet arena this year in a big way, reiterating that AMD-based Windows 8 slates are indeed en route, though she stopped short of giving an ETA. Finally, the company's renewing its focus in the server market, as it seeks to cut a larger slice of the cloud computing pie.

 

[AlexV]

Read the reports from the analyst day, especially Lisa Su's comments with regards to graphics.

That is not significantly informative though, in terms of viability of forward-looking statements. There have been a number of claims made about Bulldozer at one analyst day or another, across the years, and most fell flat on their face. As Bohr (IIRC) said, predictions are hard, especially when they involve the future.

 

[Florin]

AMD - What went wrong?

Chillblast has confirmed it currently has “zero availability” of Fusion chips, with no confirmed date for new stock. In fact, since Llano’s June launch, Chillblast has been able to secure precisely two of these chips, and Sawyer said one supplier he uses had managed to source only ten of these sought-after processors.
A spokesman for another major British firm told us off-the-record it’s had issues with both Fusion and Bulldozer, saying “we’d struggle to make a high-end gaming system” because of a lack of availability.

Perhaps the biggest threat to AMD comes from British chip designer ARM. It already has the lion’s share of the smartphone and tablet processor market; now it’s expanding into PC territory too.
Not only is Microsoft adding support for ARM into Windows 8, but the ARM-based Tegra 3 is targeting laptops. ARM’s executive vice president of marketing, Lance Howarth, has said that by 2020 “there’s only going to be two [chip designers] – ourselves and Intel”.
Morales echoes this sentiment, saying he expects AMD to become “a shell of a company” in the near future, as “tablets and smartphones move extremely fast and hardware companies can’t keep up”.

Nothing new here for industry followers but it's a decent summary of how AMD ended up in the current state.

 

[Pete]

Anything to The Verge's rumor that Eric left for Qualcomm?

 

[Dave Baumann]

http://techreport.com/articles.x/22452/2

 

[DuckThor Evil]

http://techreport.com/articles.x/22452/2

Considering that and the rumours of him going to Qualcomm, would suggest that he was disappointed with AMD's weak presence and near to mid term plans of not being in smartphones etc.?

 

[Silent_Buddha]

http://techreport.com/articles.x/22452/2

From that...

In fact, Su described graphics as one of AMD's "crown jewels" that ends up being the "centerpiece of our roadmap."

Hopefully that's more than just lip service. And until we see evidence otherwise there's no reason to not to believe what she says.

Still, it'd be awesome if Sireric suddenly popped back onto the board to give some insight into why he decided to leave AMD.

Regards,
SB

 

[upnorthsox]

So you/Su are saying that the future is practically all APU which is not a future that would include Eric Demers?

 

[3dilettante]

I would expect that it would be considered bad form for either AMD or Demers to go into the reasons for their parting of ways, beyond the boilerplate blurb that was published.

I do note that Qualcomm does have the appearances of being a profitable, stable, and well-run company, at least in comparison.
There's something to be said for trying something new, and something to be said for not going for a full decade working in a company that is not as described above.

 

[Dave Baumann]

So you/Su are saying that the future is practically all APU which is not a future that would include Eric Demers?

No, not at all. The comment was in reply to SB's:

But the world moves on. I can only hope it wasn't due to Eric thinking AMD were headed down the wrong path with regards to the GPU business unit and that AMD remains competitive there.

 

[Entropy]

So you/Su are saying that the future is practically all APU..

That future is already here.
Even in x86, only a small percentage of Intels CPU sales are of processors without a GPU. I don't know what percentage of AMDs processors are Brazos/Llano, but it should be a good part. Ergo, only a small minority of x86 processors are not APUs. Right now. (Of course, the larger cell phone/mobile market is already there.)

So there isn't much left for the future to do in this respect. The last frontier is memory bandwidth (and to some extent power draw), but there are options for that as well going forward, within a couple of process generations definitely. It's more a question of how fast Intel and AMD are prepared to move. If I were AMD, I'd be quite aggressive.

I don't know Eric D from other than his posts here, but he seemed more interested in graphics than in the x86 PC market per se. As long as he has his health and remains interested in graphics, I'd assume that he will keep making a contribution wherever he is in the industry, and good luck to him.

 

[GZ007]

Biggest problem with APU-s is that they lag behind discrete GPU-s by a year or even more with less performance. AMD already ditched 28nm GCN refresh for 7600,7500. The reason is for 100 percent APU-s laging behind GPU-s.

If they would release a pasive half 7750 with 30-40w TDP, who the hell would even try to buy their APU-s Llano was already 32nm Glofo while the rest of the discrete gpu-s still on several years old 40nm and it was already slower.

Right now APU-s TDP criples either the CPU or GPU. You gain limited upgrade or overclocking. And when u buy a discrete GPU(u want a nvidia for example) u have wasted silicon siting right there.
Not to mention fusing next bulldozer architecture with 8 percent IPC increase (will it even reach phenom 2 IPC ) doesnt give lot of a faith in trinity.

Right now it seems like nvidia wont have much of a problem against APU-s for a long time if they release 28nm discrete low end GPU-s.

 

[Tahir2]

SirEric left to join Qualcomm - wish him the best there.

It will be interesting to see where AMD is in a years time - we are all negative here but AMD stock has risen since Rory Read took the helm. Do investors and analyst know something we don't?