The AMD Execution Thread [2007 - 2017]

[AlexV]

Bad news:
http://ir.amd.com/phoenix.zhtml?c=74093&p=irol-newsArticle&ID=1712911&highlight=

"AMD Announces Preliminary Second Quarter Results
Revenue Expected to Decrease Approximately 11 Percent Sequentially; Company Effectively Manages Gross Margin "
AMDs reasoning for this:
"The lower preliminary revenue results are primarily due to business conditions that materialized late in the second quarter, specifically softer-than-expected channel sales in China and Europe as well as a weaker consumer buying environment impacting the company's Original Equipment Manufacturer (OEM) business."
Hope they can recover quickly and Trinity will make an impact at least as long as Intels 3rd-gen i3 are not on the market.

How does an outfit that large manage to fail so hard at forecasting it's QonQ dynamics? A few percentage points here and there fine, but when you miss by +10%, eyebrows should go up and some people in some offices should start feeling slightly uncomfortable.

I recall this article mentioning the increased use of statistics as one of the big new things that the new management at AMD is building (I was a bit dumbstruck at that being presented as some achievement unlocked, but I digress). I guess this points out that statistics are hard:smile:

 

[silent_guy]

How does an outfit that large manage to fail so hard at forecasting it's QonQ dynamics?

The same way a large pool of prominent economists almost without exception fail to forecast recessions, turns in business climate etc.

As a silicon company, you need to project 3 months out front what your customers will need. But in that in turn requires your customer to project what the end user will need.

I think it always will be a very fuzzy exercise that's only going to be right when there are no major turns in end consumer sentiment.

 

[Richthofen]

The same way a large pool of prominent economists almost without exception fail to forecast recessions, turns in business climate etc.

As a silicon company, you need to project 3 months out front what your customers will need. But in that in turn requires your customer to project what the end user will need.

I think it always will be a very fuzzy exercise that's only going to be right when there are no major turns in end consumer sentiment.

Of course it is tough to project or to give a guidance in the semiconductor business. However there is no other company in that business that fails so often in such a big way when looking at guidance and actual results. So from my point of view this is also related to AMDs management accross the board.

 

[Silent_Buddha]

Of course it is tough to project or to give a guidance in the semiconductor business. However there is no other company in that business that fails so often in such a big way when looking at guidance and actual results. So from my point of view this is also related to AMDs management accross the board.

There also aren't a lot of companies that have to compete against a juggernaut such as Intel on the CPU side. On the graphics front they are actually executing far better than Nvidia at the moment with similar performance at the top, and an ability to launch product fairly regularly (approx. 1 year refresh for all product segments top to bottom).

So management is at least fine in the graphics division. It could potentially be better in the CPU division, but I don't think anyone is going to do well going head to head with Intel year in and year out.

Regards,
SB

 

[Richthofen]

There also aren't a lot of companies that have to compete against a juggernaut such as Intel on the CPU side. On the graphics front they are actually executing far better than Nvidia at the moment with similar performance at the top, and an ability to launch product fairly regularly (approx. 1 year refresh for all product segments top to bottom).

So management is at least fine in the graphics division. It could potentially be better in the CPU division, but I don't think anyone is going to do well going head to head with Intel year in and year out.
Regards,
SB

Which remains to be seen once AMD and Nvidia release their numbers. Some analysts indicated that AMD might also have lost business to Nvidia. We will see once the numbers are out but to say its only because of Intel is an too easy answer for now and i would like to add that my comment was not only related to the current quater. You can look back at AMD over the past couple of years. Fails like this are a consistent thing with AMD which is why i think that they have a management problem in all divisions.

 

[Sound_Card]

I see them shrinking to graphics only not by choice, ironically in which I think it should be by choice. And Ironically they ditched a more globally recognized name (ATi) when it comes to graphics. Bring back ATi, ditch x86, go ARM. Unless in the foreseeable future, pile driver is a competitive monster that some indicate, x86 is not worth the trouble for them anymore.

If the rumors are true on having a chip (or two) in all three consoles, that should be a nice anchor to lean on for at least 5-7 years.

 

[Silent_Buddha]

Which remains to be seen once AMD and Nvidia release their numbers. Some analysts indicated that AMD might also have lost business to Nvidia. We will see once the numbers are out but to say its only because of Intel is an too easy answer for now and i would like to add that my comment was not only related to the current quater. You can look back at AMD over the past couple of years. Fails like this are a consistent thing with AMD which is why i think that they have a management problem in all divisions.

Think of it another way. ATI hit its low point just prior to AMD's takeover with the 2k series. Prior to that they were neck and neck with Nvidia. Slightly better IQ, very slightly slower.

Since the takeover, they've been trending up. Rv670, Rv770, Cypress (took the lead for >6 months), Cayman (small gap 10-15% to GF110), Tahiti (performance lead for ~3 months, ~performance parity since then).

As well since Rv770 AMD has consistently released products addressing all market segments. Compare that to Nvidia's performance since AMD took over ATI. Nvidia has been trending down with regards to execution while AMD's graphics division has been trending up with regards to execution. Execution being at least a bit more reliant on good management than product performance which would be more reliant on good engineers.

Really, there is little evidence that management on AMD's part with regards to their graphics division has been anything less than good. The only real blemish on their track record would be the perceived lack of quality in their drivers. Yet even there they are innovating at least as much as Nvidia. And both companies continue to have issues with their drivers from time to time.

Regards,
SB

 

[dkanter]

I see them shrinking to graphics only not by choice, ironically in which I think it should be by choice. And Ironically they ditched a more globally recognized name (ATi) when it comes to graphics. Bring back ATi, ditch x86, go ARM. Unless in the foreseeable future, pile driver is a competitive monster that some indicate, x86 is not worth the trouble for them anymore.

If the rumors are true on having a chip (or two) in all three consoles, that should be a nice anchor to lean on for at least 5-7 years.

That's funny. I wrote an article a while back why AMD won't do this:
http://www.realworldtech.com/amd-mobile-strategy/

AMD might play around with ARM, but giving up x86 is utter madness.

DK

 

[trinibwoy]

SB I don't know that you're looking at the right factors I determine whether they are executing well. The number of SKUs and release dates dont tell the whole story. It matters not if you have 20 new products in the market if you're getting outmaneuvered by your competition's old parts.

The bottom line is what defines good execution.

 

[ninelven]

AMD might play around with ARM, but giving up x86 is utter madness.

While I don't see AMD giving up on x86, I do think ARM offers them more opportunity long-term. As it is now, they are not competing with Intel, and quite frankly it is hard to believe they ever will again at this point. Perhaps they can catch up when the industry eventually moves away from silicon, but who knows what things will look like by then... I also think power efficiency is going to become a monumental issue in the next few years, even to the point it will eventually eclipse all other considerations.

Overall, I expect slow but steady growth for ARM matched by a slow and steady decline by x86. But predicting the future is hard...

 

[entity279]

Overall, I expect slow but steady growth for ARM matched by a slow and steady decline by x86. But predicting the future is hard...

But the market will shrink (!) so there shouldn't be a match between the growth and decline

 

[ninelven]

What market?

 

[Silent_Buddha]

I'm assuming he means the desktop PC market. Then again in the late 90's everyone was saying the Desktop PC market was going to shrink and be irrelevant for most consumers. And that never happened. Again in the early 2000's the same thing was said. And never happened. Ditto the late 2000s. And again now, people are saying the same thing again, and I'm going to bet that once again it doesn't happen.

Regards,
SB

 

[ninelven]

Yeah, I doubt the PC market will drastically shrink in absolute terms. It's % market share compared to "post-pc" devices will certainly continue to decline, but I would expect the number of active PCs to continue growing.

 

[Silent_Buddha]

Yeah. I have no doubt that the slate/tablet market as well as smartphone market will continue to grow. But I doubt that it will be at the expense of the PC market. They are both complimentary markets to each other with some overlap but a lot of area's where there is no overlap.

Regards,
SB

 

[sheepdogexpress]

Yeah. I have no doubt that the slate/tablet market as well as smartphone market will continue to grow. But I doubt that it will be at the expense of the PC market. They are both complimentary markets to each other with some overlap but a lot of area's where there is no overlap.

Regards,
SB

The rise of the tablets will definitely hurt AMD. Netbook and brazos laptops sales are being eaten up by tablet sales. The netbook was a solid area for growth and revenue that AMD could have used as a income source from products like brazos.. It was a market it could compete with Intel on superior footing since ATOM sucked so much. That market is disappearing overnight because of tablets. AMD is stuck selling 600-700 dollar laptops because at 800 dollars, AMD is running into to many Intel based laptops with discrete graphics.

This generation AMD seemed to drop the ball with laptop manufacturers as it seems like there are far less 7xxx based laptops vs 6xxx in regard to discrete graphics.

With the netbook market fading, AMD is very niche now because the only market it can hope to compete, especially against the intel brand is the core i3 market with only integrated graphics. The public prefer Intel branding too much. And a heavier emphasis is put on the CPU compared to the GPU.

 

[CarstenS]

Did no one pick up the Q2 number from AMD yet?

A little bit surprising, but surely very welcome they managed to produce a net-income of about 37 million, despite revenue declining about 10% yoy (11% qoq).

Good job in a difficult environment!

More:
http://seekingalpha.com/article/734...-results-earnings-call-transcript?part=single

 

[trinibwoy]

. That market is disappearing overnight because of tablets.

As well it should. Netbooks were one of the worst ideas ever conceived.

 

[sheepdogexpress]

Did no one pick up the Q2 number from AMD yet?

A little bit surprising, but surely very welcome they managed to produce a net-income of about 37 million, despite revenue declining about 10% yoy (11% qoq).

Good job in a difficult environment!

More:
http://seekingalpha.com/article/734...-results-earnings-call-transcript?part=single

Much of this has to do with losing 10 percent of the workforce. Which according to reports would save 118million dollars for AMD in 2012. In addition canceling project like witchita and Krishna would have also saved money.

In the long run I don't think think saving the money was worth it because they let go of people that were to crucial in AMD CPU division. Brad Burgess the guy responsible for Brazos and it's successor and John Bruno, the guy responsible for trinity. The cuts were too broad and AMD should have not cut guys who were responsible for many of the successes with their CPU division in the last couple years.

 

[Kaotik]

Wichita & Krishna cancellations costed AMD if anything, as they were pretty much "good to go" excluding the fact that the GF foundries weren't up to the task, thus a lot of possible profits were lost.