09-27-2013, 04:27 PM
(This post was last modified: 09-27-2013, 04:29 PM by hendi_alex.)
Matt who posts on my Income Investing and Beyond Board at Investor Village seems to be pretty savvy wrt to tech. He has had some well thought out posts related to both Apple and Intel. I don't think that he would mind me sharing this one in depth INTC post. I'm overweight INTC at 13% weighting but also have sold calls against 100% of the position. Have generally been long INTC for most of the past three years except for a few months when all shares got called. Cuurently my calls are laddered from $21 up to $24 and my average cost is a bit over $22. Dates go out as far as April.
Well, that's a layman's view, but it isn't what is really going on. Basically the media has hyped up the mobile/pad thing to the point where it is attributing the shift well past the markets where it actually works. A lot of us engineers have been scratching our heads over all the ridiculous ARM hype.
Basically it comes down to this: There is a major shift away from laptops and PC's in the home consumer market and towards mobile devices (pads and phones). It's a big market, which is why many companies (Dell, HP) are feeling the pain. ARM is competitive in this market primarily because ARM focuses on drawing very little power when the cpu is mostly idle. The shift is occurring because laptops and desktops are far more powerful than consumers really need. They're happy with pads and other mobile devices despite those devices being 5x to 10x less capable.
But that is about as far as it goes. Intel has the server space locked up. ARM has exactly zero chance of penetrating it. Servers don't need low power low-performance cpus. Servers need modestly low power HIGH performance cpus. ARM can't even remotely compete with Intel in the Server market. It isn't even close, and that's comparing ARM's not-yet-in-production technology against Intel's 2-year-old technology. When you throw Haswell into the mix Intel has at least a 5-year lead on everyone else which is insurmountable.
In the business laptop and desktop space ARM also has basically no chance. These devices need to perform, or employee productivity suffers. Should be obvious to anyone, frankly. Try writing a few pages of text on a mobile device.... painful as hell, even with a bluetooth keyboard. Similarly, employees with business laptops often leave many apps open and active at once, have browsers with 50 tabs open, and expect it all to be zippy. Mobile cpus have basically no chance of being able to handle that sort of workload. Throw in Haswell laptops and nobody is going to be buying ARM devices for business laptop needs.
Even worse, Haswell is likely going to blow AMD away in the laptop market. So while Intel will lose share to pads on the consumer side, the losses will be mitigated by AMD losing pretty much the remaining small percentage of desktop and laptop market that it had to Intel. AMD can't even remotely compete with Haswell either.
--
So essentially we are in a situation where the markets believe Intel's base business will continue to suffer severe setbacks whereas I believe that much of that is now baked in... the consumer space is as switched over to mobile as its gonna get relative to consumer laptop and desktop sales, and Haswell is virtually certain to trigger a new business laptop/desktop cycle as well as a new server cycle. Literally Haswell pays for itself in the server environment in just a few months... that's how important it is in that space.
Notice that I haven't even talked about Intel penetrating the mobile market yet. Basically, so far, I've laid the case down for Intel's existing business lines stabilizing.
Now throw in Intel's penetration of the mobile market and keep in mind that Android (in particular) runs just fine on Intel cpus, and that mobile phone makers basically do an entire hardware reorg every 8 months. What does this mean for Intel? It means that Intel only has to produce something which is on-par with ARM's offerings, what is called 'System on a Chip', for mobile, to be able to penetrate the market. ARM has no protective moat. And mobile is far larger than the consumer space ever was. Once Intel begins to penetrate the mobile market it will have no problem making up for loss in the consumer desktop and laptop markets, and it will STILL have its business and server markets held totally intact.
Now throw in Intel's FAB advantage. All Intel needs to compete with ARM is technology that is already about to be released late this year / early next year (I am discounting the work-up technology they've already released in the mobile space). Intel has two more generations in the pipeline after that. At a minimum that means that Intel will be on-par with any future advances ARM and ARM vendors can make. In otherwords, within two years Intel is going to have a major chunk of the mobile cpu market no matter what happens in the ARM space. That's the absolute worst case.
When you look at power use and related considerations, ARM's only real advantage was its lower-power consumption AT IDLE. Intel is already competitive with ARM's power consumption just using Intel's work-up mobile technology. Within one generation of its current technology Intel will be handily beating ARM on power consumption for mobile devices. And I should add that when the cpu is loaded, ARM is already non-competitive with Intel. So when it comes to requiring more performance out of mobile devices Intel is already in very good shape with current technology.
In anycase, that is what is really going on in the space.
-Matt
posted 8/16/2013
Well, that's a layman's view, but it isn't what is really going on. Basically the media has hyped up the mobile/pad thing to the point where it is attributing the shift well past the markets where it actually works. A lot of us engineers have been scratching our heads over all the ridiculous ARM hype.
Basically it comes down to this: There is a major shift away from laptops and PC's in the home consumer market and towards mobile devices (pads and phones). It's a big market, which is why many companies (Dell, HP) are feeling the pain. ARM is competitive in this market primarily because ARM focuses on drawing very little power when the cpu is mostly idle. The shift is occurring because laptops and desktops are far more powerful than consumers really need. They're happy with pads and other mobile devices despite those devices being 5x to 10x less capable.
But that is about as far as it goes. Intel has the server space locked up. ARM has exactly zero chance of penetrating it. Servers don't need low power low-performance cpus. Servers need modestly low power HIGH performance cpus. ARM can't even remotely compete with Intel in the Server market. It isn't even close, and that's comparing ARM's not-yet-in-production technology against Intel's 2-year-old technology. When you throw Haswell into the mix Intel has at least a 5-year lead on everyone else which is insurmountable.
In the business laptop and desktop space ARM also has basically no chance. These devices need to perform, or employee productivity suffers. Should be obvious to anyone, frankly. Try writing a few pages of text on a mobile device.... painful as hell, even with a bluetooth keyboard. Similarly, employees with business laptops often leave many apps open and active at once, have browsers with 50 tabs open, and expect it all to be zippy. Mobile cpus have basically no chance of being able to handle that sort of workload. Throw in Haswell laptops and nobody is going to be buying ARM devices for business laptop needs.
Even worse, Haswell is likely going to blow AMD away in the laptop market. So while Intel will lose share to pads on the consumer side, the losses will be mitigated by AMD losing pretty much the remaining small percentage of desktop and laptop market that it had to Intel. AMD can't even remotely compete with Haswell either.
--
So essentially we are in a situation where the markets believe Intel's base business will continue to suffer severe setbacks whereas I believe that much of that is now baked in... the consumer space is as switched over to mobile as its gonna get relative to consumer laptop and desktop sales, and Haswell is virtually certain to trigger a new business laptop/desktop cycle as well as a new server cycle. Literally Haswell pays for itself in the server environment in just a few months... that's how important it is in that space.
Notice that I haven't even talked about Intel penetrating the mobile market yet. Basically, so far, I've laid the case down for Intel's existing business lines stabilizing.
Now throw in Intel's penetration of the mobile market and keep in mind that Android (in particular) runs just fine on Intel cpus, and that mobile phone makers basically do an entire hardware reorg every 8 months. What does this mean for Intel? It means that Intel only has to produce something which is on-par with ARM's offerings, what is called 'System on a Chip', for mobile, to be able to penetrate the market. ARM has no protective moat. And mobile is far larger than the consumer space ever was. Once Intel begins to penetrate the mobile market it will have no problem making up for loss in the consumer desktop and laptop markets, and it will STILL have its business and server markets held totally intact.
Now throw in Intel's FAB advantage. All Intel needs to compete with ARM is technology that is already about to be released late this year / early next year (I am discounting the work-up technology they've already released in the mobile space). Intel has two more generations in the pipeline after that. At a minimum that means that Intel will be on-par with any future advances ARM and ARM vendors can make. In otherwords, within two years Intel is going to have a major chunk of the mobile cpu market no matter what happens in the ARM space. That's the absolute worst case.
When you look at power use and related considerations, ARM's only real advantage was its lower-power consumption AT IDLE. Intel is already competitive with ARM's power consumption just using Intel's work-up mobile technology. Within one generation of its current technology Intel will be handily beating ARM on power consumption for mobile devices. And I should add that when the cpu is loaded, ARM is already non-competitive with Intel. So when it comes to requiring more performance out of mobile devices Intel is already in very good shape with current technology.
In anycase, that is what is really going on in the space.
-Matt
posted 8/16/2013
Alex