Much chatter today about whether Apple is going to drop Intel and x86 for the Mac line at some point and move over to something ARM based. Everyone’s agreed that it’s not going to happen immediately, but perhaps in four or five years?
Apple Inc. (AAPL) is exploring ways to replace Intel Corp. (INTC) processors in its Mac personal computers with a version of the chip technology it uses in the iPhone and iPad, according to people familiar with the company’s research.
Apple engineers have grown confident that the chip designs used for its mobile devices will one day be powerful enough to run its desktops and laptops, said three people with knowledge of the work, who asked to remain anonymous because the plans are confidential. Apple began using Intel chips for Macs in 2005.
After all, it’s not as if Apple hasn’t made such architecture transitions before:
While a move away from Intel would be a massive undertaking, it’s a transition that Apple has experience with â€" its move from PowerPC processors to Intel went about as smoothly as anyone could have expected, though support for older machines and software was eventually cut off. At the time, it was revealed that Apple had been secretly building versions of OS X to run on Intel hardware for nearly all of the OS’s development lifecycle â€" and it’s entirely possible that Apple is similarly building versions of its desktop software to run on future, Apple-designed chipsets.
And don’t forget they made the move from 68x chips to Power PC before that too. It’s certainly considered technically feasible:
A quick look at the latest Apple silicon shows the company is on the right track. The A6X is a serious piece of silicon that makes the newest gen 4 iPad a powerful gaming tablet. Just check out these benchmarks at Anandtech.
If Apple maintains this pace, there’s no reason it couldn’t put, let’s say, a future A7 or A8 in a MacBook-like device.
I think the important point comes here:
But since that 2005 tie-up with Intel, Apple acquired P.A. Semi and turned its technology into the Ax microprocessor range. Those chips now ship in the tens of millions, thanks to their presence in iPhones and iPads, products that dwarf Apple’s iMac business.
The ARM chips are great for power management, this is true. But they do rather lack the graphics and number crunching ability of the Intel chipsets. However, I have a very strong feeling that as a box maker, a hardware manufacturer, you want to have your technology based on whichever chipset is being developed most widely. For the more people using it, the more people designing using the basics, then the faster that chipset is likely to advance in its capabilities.
This isn’t to say that Intel are bad at development, or sluggards at it, of course not. But ARM in the UK really only design the very core. There are multiple firms across the world adding to that core design in a manner that just doesn’t happen in the Intel ecosystem. Further, there are multiple fabs around the world that can manufacture the designs once they’re complete. Just as it isn’t true that the race is always to the swift, not the battle to the strong but that’s still the way to bet, I would assume that a chip ecosystem with multiple designers and multiple manufacturers is going to advance more swiftly than one being developed by only one company.
In the longer term I would expect ARM to be both more capable and also more adjustable to a specific hardware manufacturer’s desires than the Intel designs. Which is why moving from Intel to ARM may well be a very good idea.
Not to mention the fact that if there are multiple potential manufacturers you can squeeze them on their profit margins in a way that you cannot Intel….
No comments:
Post a Comment