Sunday, June 21, 2020

ARM Mac Impact on Intel

by Jean-Louis Gassée

Apple Macs are only a small part of the least profitable segment of Intel’s CPU business. On the Surface of things, nothing for Intel to worry about…today. But, in the long run, Microsoft and Apple might become unintentional co-conspirators against Intel’s x86 business.

According to a recent Statista tally, Mac CPUs took about 12% of the US PC market. NET MARKETSHARE pegs it slightly above 9%; Gartner says it’s now 6.9%. Furthermore, as Intel’s latest quarterly numbers show, the business segment that supplies high-end chips to data center servers has higher operating margins (50%) and grows faster (+43% compared to the same period a year ago) than its conventional PC segment (36% and +14% respectively). Intel’s real money is in high-end CPUs sold to prosperous Cloud operators, not in supplying lower-end chips to cost-cutting laptop makers.

Let’s not meditate too long on the value and biases in these numbers and just accept that Intel won’t shed many tears over the loss of Apple’s small share of a lower margin business. But has the current focus on Apple’s presumptive repudiation of x86 CPU chips blinded us to a more momentous consequence for Intel?

At the January 2011 CES, Microsoft announced a version of Windows running on ARM processors. On February 9th, 2012, Windows President Steven Sinofsky released an authoritative post, titled Building Windows for the ARM processor architecture, that gave further details of the new platform. Four months later, Microsoft introduced the Surface, a PC-tablet convertible built around an ARM processor and running Windows on ARM (WoA), soon renamed Windows RT.

The Surface device, a first for Microsoft, was positioned as a Design Point, a sort of inspiration for PC makers rather than a replacement. Demos to selected members of the press resulted in favorable reviews, such as Why I Love Surface from Farshad Manjoo [as always edits and emphasis mine]:

“I’m already deeply smitten. Not because the Surface is so great — though it seems like it might be — but because it represents a new and potentially powerful force in the tech industry. For the first time in its history, Microsoft is taking PC hardware as seriously as it does software. The software giant is coming around to a maxim that archrival Steve Jobs always held dear — that the best technologies come about from the tight integration of code and manufacturing, and that no company can afford to focus on just one half of that equation.”

At the time, the move made even more noise than today’s ARM-ing of the Mac. Imagine the industry upheaval, the impact on Intel if PC OEMs followed Microsoft’s Design Point and built personal computers based on ARM Architecture processors. And imagine the conversations between Microsoft execs and key PC makers — and the consternation inside Intel. The Wintel duopoly would dissolve, replaced by a “proper” Microsoft stewardship of the PC business and a “proper” competitive world of ARM Architecture processor suppliers. Smarter and less expensive Windows personal computers.

As we know, nothing of the sort happened.

Impressed by the initial Surface reviews, I immediately bought one. But after I experienced the actual product, I was only too happy to resell the device to a developer who was eager to bring one back to Eastern Europe. I wasn’t alone: Once units got in the field, mixed — or frankly bad — reviews followed and sales were worse than modest. In 2013, Microsoft took a $900M loss on the project. The company released a couple of iterations but for the next few years we didn’t hear much about Windows running on ARM devices.

Nonetheless, Microsoft continued to work on an ARM future for Windows.

Late 2019, the company introduced the Surface Pro X, an elegant hardware device in the tradition that Microsoft had established for its Surface brand. Unlike current Surface laptops and hybrid PC-tablets, which all run on x86 Windows, the Surface Pro X was another go at introducing the ARM wolf into the PC sheepfold. Unfortunately, the reviews haven’t been generous, with titles such as This isn’t the long-lasting tablet we were hoping for, Microsoft Surface Pro X review: not yet ready for prime time, or Surface Pro X Tries Again, Fails Again:

“Microsoft has ported Windows 10 to ARM, so the operating system runs natively. But the company’s premier business apps, like Word, Excel, PowerPoint, the new Edge browser, and Microsoft Teams (its alternative to Slack) haven’t been ported; they all run x86 binaries in emulation. In fact, very little Microsoft software runs natively on Microsoft’s new flagship platform. Odd.

But the situation is far worse for third-party apps, many of which flat don’t work. On Surface Pro X, third-party software falls into four categories: some apps run fine, some run slowly, some install but don’t work, and some won’t even install. Sadly, there’s no way to tell which apps fall into which category, and this may be the Pro X’s biggest failing.”

This is puzzling. One assumes Microsoft’s top executives were aware of the inadequacies, but they approved the Surface Pro X launch anyway. Done well, the new device could have been a signal to PC OEMs: “This time we got it right. Get ready for a new era of better laptops and hybrid PC-tablets, lighter, less expensive to build, Intel-free!” Instead, Intel execs must have smiled discreetly as Lenovo, HP, and Dell product planners returned to more-of-the-same design sessions.

A few thoughts when contemplating Microsoft’s sorry attempts at freeing Windows PCs from Intel and Apple’s proposed move.

First, we’ll see how Apple manages to succeed where Microsoft failed…twice. Yes, Apple knows how to port its devices to a new CPU. As mentioned in last week’s Monday Note, the move from PowerPC to Intel in 2005 was an unqualified success. But will Apple’s inducement to third party app developers — the (retroactively understandable) rejection of 32-bit apps in macOS Catalina, the Catalyst tool that will help developers bring their iOS apps to macOS — will the groundwork pay off?

Second, Microsoft execs might be thinking longer term than today’s “try again, fail again” ARM efforts. Perhaps they’re thinking that by the time Apple ships its ARM-ed Macs, the next generation Surface Pro X will enjoy a full suite of ARM-native apps from Microsoft and key third parties. This would result in a product that’s competitive with whatever Apple has next year.

Finally — and more relevant to today’s topic — maybe it’s not a zero-sum, Apple vs Microsoft game. Fully functional ARMed Windows and macOS devices might finally have an impact on Intel’s business. If both Microsoft and Apple successfully unleash attractive ARM Architecture devices, Windows OEMs big and small won’t want to be left behind; they’ll launch a plethora of machines running the next generation of ARM-based software. That might finally have a sizable impact on Intel’s business.

Now, imagine a world where Windows machines powered by x86 and ARM Architecture CPUs coexist…

We won’t be bored.

PS: For today, we’ll leave the impact of ARM-based servers and their greater thermal efficiency alone.

JLG@mondaynote.com



from Hacker News https://ift.tt/2NnCjAF

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.