It's an situation with lots of potential for Apple - they *could* turn things on their head - they provide the platform, manufacturers come to Apple to make the hardware accessories - not the other way round. You can sell one laptop, one iPhone, but several accessories to a person.
I think it's not going to get into embedded devices for others - that work is for it's own Macs, and other Apple products
However, Apple's going to make it easier for companies to create an Apple friendly embedded device work with an iPod and iPhone.
Working to get companies to bring the integration *to them*
Apple works on the interface, and dangles the App Store and iPhone at companies - we'll see who comes running - likely those who have less to lose from such integration with another company.
Work with maybe the car industry a little. Get the uptake for the iPhone as a SatNav and a music player in the car - replace CDs (and if you have signal, replace radio).
I agree about radio - it's going to change. I think Apple held back for a year, as they knew that they could get convergence the SatNav market once they had turn by turn out - (OS 3.0)
But what do they have now and coming? Bluetooth 2.1/3, 802.11n and more. they're opening up the dock - There's a big incentive for the first company to get integration done well
- e.g. the first CD manufacturer to get good linking to an iPhone or Touch
- e.g. The first sport equipment producer to be able to create retrofit kits, or new kit with iPhone docks/cables/BT etc
(Imagine if you could get your workout data onto an app after going to the gym - sync that row on the Concept II ergo - check the HR on the middle section, upload the 10 miles on the treadmill to Nike Plus website etc)
Chip design
It links in Intel, Nvidia, GPU and CPU, x86 and ARM, PA Semi Imagination Technologies and more. Dependent on how fast Apple can optimise for massive parallelism they could take multi-core, and massively-multi-core to both Macs (desktop, laptop) and their MIDs through to their iPods.
A chip design could be utilised in an entire product family - look at how the 9400M spans iMac, Mac mini, MacBook, and MacBook Pro. How much work they can get with their main CPU and GPU providers is unknown. Maybe they've been able to cosy up to Nvidia a bit recently, and have got some interesting new kit readying in the next year or so. Time will tell.
As chewietobbacca points out - Apple does little production. But then the rumors aren't about actually owning a fab - but having a hand in design, and integration to their hardware. If they can work with partners on adding modular components to create fewer chips on a board, then that's a win.
Apple's taken on PA Semi who do actual design of components. They've likely got an ARM license. They won't be making things from scratch - but surely they'll have enough human resources to better put things together?
An artist doesn't usually make their paints. But they sure do mix the colors to their preference, then paint them out as a bad analogy.
And what's so wrong with not going x86 with the CPU or GPU? It's hard for Apple - Intel's messing people around - they want to control a market they aren't good at (things including GPU) whilst they control CPU. Apple likes them for CPU - but would presumably want GPU from elsewhere, and the increasing power of ARM does give pause for breath. If Intel could get GPU right, wouldn't Intel grab the opportunity to get it, if the TDP , power consumption and price were right? They are a giant. Pissing them off presumably wouldn't serve Apple well at all. But the convergence of GPU and CPU isn't going to go away. It puts Apple between 2 companies to an extent.
If Intel and AMD and ATI and Nvidia plan their CPUs/GPUs so far ahead, then does this mean Apple's decided that far ahead as well?
Even small scale differentiation, if it leverages their strengths, will do wonders. There likely won't be anything *big* on the industry-shattering scale. But then we do have a one of leap in potential computing power from incorporating GPU into the computer as a GPGPU.
Several years after this event, there won't be such a change year on year- it'll be more steady progress. But at this time, it seems that there is a large, one-off performance boost to be had from changing the concept of what provides the grunt power to a computer and getting the GPU on board.
I think it's not going to get into embedded devices for others - that work is for it's own Macs, and other Apple products
However, Apple's going to make it easier for companies to create an Apple friendly embedded device work with an iPod and iPhone.
Working to get companies to bring the integration *to them*
Apple works on the interface, and dangles the App Store and iPhone at companies - we'll see who comes running - likely those who have less to lose from such integration with another company.
Work with maybe the car industry a little. Get the uptake for the iPhone as a SatNav and a music player in the car - replace CDs (and if you have signal, replace radio).
I agree about radio - it's going to change. I think Apple held back for a year, as they knew that they could get convergence the SatNav market once they had turn by turn out - (OS 3.0)
But what do they have now and coming? Bluetooth 2.1/3, 802.11n and more. they're opening up the dock - There's a big incentive for the first company to get integration done well
- e.g. the first CD manufacturer to get good linking to an iPhone or Touch
- e.g. The first sport equipment producer to be able to create retrofit kits, or new kit with iPhone docks/cables/BT etc
(Imagine if you could get your workout data onto an app after going to the gym - sync that row on the Concept II ergo - check the HR on the middle section, upload the 10 miles on the treadmill to Nike Plus website etc)
Chip design
It links in Intel, Nvidia, GPU and CPU, x86 and ARM, PA Semi Imagination Technologies and more. Dependent on how fast Apple can optimise for massive parallelism they could take multi-core, and massively-multi-core to both Macs (desktop, laptop) and their MIDs through to their iPods.
A chip design could be utilised in an entire product family - look at how the 9400M spans iMac, Mac mini, MacBook, and MacBook Pro. How much work they can get with their main CPU and GPU providers is unknown. Maybe they've been able to cosy up to Nvidia a bit recently, and have got some interesting new kit readying in the next year or so. Time will tell.
As chewietobbacca points out - Apple does little production. But then the rumors aren't about actually owning a fab - but having a hand in design, and integration to their hardware. If they can work with partners on adding modular components to create fewer chips on a board, then that's a win.
Apple's taken on PA Semi who do actual design of components. They've likely got an ARM license. They won't be making things from scratch - but surely they'll have enough human resources to better put things together?
An artist doesn't usually make their paints. But they sure do mix the colors to their preference, then paint them out as a bad analogy.
And what's so wrong with not going x86 with the CPU or GPU? It's hard for Apple - Intel's messing people around - they want to control a market they aren't good at (things including GPU) whilst they control CPU. Apple likes them for CPU - but would presumably want GPU from elsewhere, and the increasing power of ARM does give pause for breath. If Intel could get GPU right, wouldn't Intel grab the opportunity to get it, if the TDP , power consumption and price were right? They are a giant. Pissing them off presumably wouldn't serve Apple well at all. But the convergence of GPU and CPU isn't going to go away. It puts Apple between 2 companies to an extent.
If Intel and AMD and ATI and Nvidia plan their CPUs/GPUs so far ahead, then does this mean Apple's decided that far ahead as well?
Even small scale differentiation, if it leverages their strengths, will do wonders. There likely won't be anything *big* on the industry-shattering scale. But then we do have a one of leap in potential computing power from incorporating GPU into the computer as a GPGPU.
Several years after this event, there won't be such a change year on year- it'll be more steady progress. But at this time, it seems that there is a large, one-off performance boost to be had from changing the concept of what provides the grunt power to a computer and getting the GPU on board.