Apple's announced that it is switching processors from Intel to ARM. (For the MacBook. Apple has used ARM for iPhone and iPad for years.) This announcement indicates problems for Intel.
But first, some thoughts on "market leaders".
We like to think that markets have leaders, and those leaders set the direction and pace for the market, much as a squad leader in the military sets the direction and pace for the squad. Let's call this the "true leader".
There is another type of leader, one that instead of setting the direction and pace, looks at the crowd and runs out in front, and then claims leadership. Let's call this the "opportunistic leader".
So is Intel a true leader or an opportunistic leader? I think it is the latter.
One could argue that Intel lost its ability to lead the market in the mid-1980s, with the specific culprit being the IBM PC. Prior to the IBM PC, the market for personal computers was fractured, with different manufacturers using different processors. Apple and Commodore used the 6502, Tandy used the Zilog Z-80 in the TRS-80, and lots of others used Intel's 8080. Intel had the 8088 and 8086 processors, but very few manufacturers used them.
In the 1980s, Intel had a plan for the future of microprocessors - the iAPX line. It was a plan that built on the 16-bit 8086 but expanded to different architectures as the processors became more powerful. There were several specifications for processors, culminating with the iAPX 432 a 32-bit processor. (At the time, this was impressive.) The more powerful processors were not compatible with the 8-bit 8080, or (notably) the 16-bit 8086. It was a nice plan.
The IBM PC and the market for the IBM PC invalidated that plan. People wanted the IBM PC, which meant the Intel 8088. Later, it meant the IBM PC AT with the Intel 80286, which was not an iAPX processor.
This shows that Intel was not leading the market in the "true leader" sense.
Want another example? Let's look at "itanium", Intel's attempt at 64-bit architecture. (Also known as "IA-64".) It was a processor that was not compatible with the existing 32-bit software. Worse, it has a design that pushed work to the compiler, and some of that work was not determinable at compile time. AMD countered with x86_64, a design compatible with existing 32-bit code, and customers wanted that instead of Intel's offering. They wanted it so much that Intel eventually adopted the x86_64 design and abandoned IA-64.
I think these two examples show that Intel is more of the opportunistic leader than the true leader. Intel can design processors, and can design and manufacture the chips and build the support circuitry, but it's business strategy is to run out in front of the market.
Which brings us to Apple and ARM processors.
Apple has announced that it will switch from Intel processors to ARM processors for its Macbook laptops.
Is Apple a true market leader? Or are they opportunistic, running out in front of the crowd?
It doesn't matter.
It may be that Apple's decision will "move the market". It may be that Apple's "spidey sense" has detected a desire to shift to ARM processors, and Apple is first in the market to make the change. Whichever reason, Apple is doing it.
I think that not only Apple will shift from Intel to ARM, but the rest of the market will, too. For two reasons: cost and efficiency.
People will want computers with ARM processors because they (the computers) will have a lower price.
People will want computers with ARM processors because they (the computers) will be more energy-efficient. (Which means a longer battery life.)
People who buy computers generally don't care about the processor. They care about running applications, and ARM can deliver that.
Windows is ready to move to ARM. Linux is ready to move to ARM. Apple is not only ready to move to ARM but is in fact moving to ARM.
What will happen to Intel?
Intel has a definite problem. They have lost the low-end market for processors. They are losing the mid-range market for processors.
One might think that Intel may find haven in high-end processors, running servers and other high-power computers. And there may be some space for them, but there is a problem: IBM.
IBM has staked out the top end of the market, with System/z processors (the successor to IBM's System/360, System/370, and System/390 processors) running the big servers that host virtual machines. IBM has a competent business there, one that will not be easily displaced by Intel's x86 processors.
Intel has a market that it being eaten by ARM at the bottom end and limited by IBM at the top end. Not an enviable position.
And certainly not a market leader, in any sense.
No comments:
Post a Comment