A key reason why the pace of Mac hardware updates has slowed in recent years can be placed in the hands of Intel. As they fell behind releasing new silicon, Apple was left without the parts they needed to do proper refreshes. True, Apple did allow the Mac to languish way beyond the release of new Intel parts.

Now consider the 2016 MacBook Pro. The critics pounced on Apple’s perceived lapses. So why didn’t it support 32GB RAM, even though no previous Mac notebook supported more than 16GB? Even the Microsoft Surface, with which they are compared, doesn’t have that support.

So I won’t get into the oft-repeated reasons. Let’s just say that a future Intel chipset might well provide such support without compromises, so maybe it’ll happen with the 2018 MacBook Pros.

The other argument was that Apple failed to use the latest Intel chips, known as Kaby Lake. Once again, some critics overlooked the fact that they weren’t shipping in quad-core versions last fall; they are now. So the MacBook, MacBook Pro and iMac received appropriate updates last week.

You might have seen the early benchmarks. The performance boost of the new gear is in the low double digits. It may not sound like much, but after single digit increases in recent years, it’s a revelation. Macworld reported Geekbench results that reveal a single core improvement of 16%, and a multicore improvement of 19%.

That takes us to this year’s iPad Pro refresh, with Apple’s A10X Fusion chip. It’s an interesting design, with six cores, three of which operate at high power, and three of which operate at low power to improve battery life. The GPU has 12 cores.

But none of that matters, except how it impacts performance. Here, performance for single core tasks is 30% faster than the previous version, according to Apple. For multicore tasks, it’s 82% faster.

More to the point, Apple claims that the new iPads are “more powerful than most PC laptops.” Intel take note!

Forget about other ARM-based tablets from Samsung and other companies. If you compare the benchmarks for a 2017 iPad Pro against a MacBook, it’s a lot faster. Compared to last year’s MacBook Pro, it comes close in single core tests, although multicore tests reveal a wider performance gulf.

But you have to put some conditions on such results. One article I read that includes such a comparison didn’t bother to mention how much faster the 2017 MacBook Pro runs. But consider that Intel has been building microprocessors since 1971. Although AMD tries to be the scrappy competitor, a healthy majority of personal computers still use Intel silicon. While Apple has used other chips throughout its history, it joined the crowd in 2006 with the very first Intel-based Macs.

In contrast, Apple’s first in-house chip, the A4, based on the ARM architecture, came out in 2010. So Apple has been releasing A-series chips for seven years, and is now building parts that provide performance that exceeds or competes with notebook computers with Intel inside. The rapid run-up in power is a reason why some suggest that Apple ditch Intel and put ARM inside Macs.

Now there are reasons why what sounds good on paper may not sound good when you look at the whole picture. Even if Apple makes it easy to port Intel apps to ARM, there will have to be a period where older apps run in emulation; in other words slower. One key advantage of Apple moving to Intel was to allow you to run Windows on a Mac at full speed in Boot Camp, and with pretty decent performance via a virtual machine. A lot of people use Macs as dual-OS machines; I do.

Now Apple may find a way to make its A-series chips powerful enough to emulate Intel without a noticeable speed drop. Maybe you’ll be able to run Windows that way too. Migrating from one processor family to another will entail a fair amount of pain, even if you can do the basics with a little pointing and clicking in Xcode. Developers will still have to do a fair amount of compatibility and performance tuning to make their apps run at full efficiency.

Apple says it won’t happen. So it appears that the closest you’ll get will be the use of A-series chips for low-power functions, such as the MacBook Pro’s Touch Bar. There are reports that the Power Nap feature may also get similar treatment, but it’ll never the whole widget.

Still, the fact that Apple can beat most mobile chips from Intel in seven short years has to be a miracle. Sure, Apple has the advantage of being able to optimize its silicon to meet the needs of just three operating systems — iOS, watchOS and tvOS. The number of models supported is also relatively small. This allows Apple to take shortcuts, and not worry about functions it doesn’t need.

In contrast, Intel needs to meet the needs of numerous manufacturers with thousands and thousands of different configurations. Embedded in those Core and Xeon chips are all sorts of legacy functions that have to  accounted for. No doubt having many masters means less efficiency.

But if you consider the pace of growth, it won’t be long before Apple’s chips speed way beyond the best Intel has to offer. Even if the Mac doesn’t make another processor switch, a future iPad or a convergence computer of some sort may do things far quicker than you might expect. Do you remember how slow those original iPads were?

This article was posted on Wednesday, June 14th, 2017 at 12:00 AM and is filed under News and tagged with: A10X Fusion, A4, AMD, Apple, ARM, Boot Camp, Core, Intel, iPad, iPad Pro, Kaby Lake, Mac, Macbook, Macbook Pro, Microsoft Surface, Touch Bar, Virtual Machine, Windows, Xeon.