What If Snapdragon X Marks the Spot … for Chromebook? (Premium)

It’s probably not surprising that my recent MacBook Air M3 triggered a rabbit hole of conjecture about coming Snapdragon X-based Windows laptops. But what about ChromeOS? Isn’t this lightweight personal computing platform an even better match for this intriguing new family of Arm chips?

Don’t get me wrong, I do believe that Snapdragon X could transform Windows on Arm from an also-ran into a mainstream solution and provide the foundation for a new decade of PC use. But I also have concerns and fears based on too many years of experience and too many disappointments. And while Snapdragon X seems to address the hardware-based disappointments of the past, there are still some concerns on the software side of this equation.

These concerns are part of a much broader topic, a topic I’ve been grappling with since the web and then mobile devices transformed the industry. Many years ago, I voiced these concerns in the form of a semi-rhetorical question about which outcome was more likely: Microsoft simplifying Windows, its complex legacy platform, or Apple (at the time) making its simpler iOS mobile platform more sophisticated?

The history since is messy. Microsoft has tried to simplify Windows in more ways than most remember, with Windows RT, Windows 10X, S mode and Windows 10 S, Windows 11 (albeit all surface-level), and so on. Apple has steadily improved iOS, splitting off iPadOS as a semi-independent fork, and adding laptop capabilities. And ChromeOS, which started off as the simplest-possible platform, has expanded with Android and Linux app compatibility and more premium hardware experiences. No clear winner has emerged. Today, Windows PCs, iPads, and Chromebooks are all viable product lines, but none has replaced the others. We are, in many ways, in the same place we were 10 or 15 years ago or more.

I feel strongly that simplicity wins in the end, but I sometimes have to remind myself that those simpler platforms—iPad and Chromebook—are offshoots of far more complex, legacy platforms too. The iPad came out of the iPhone, obviously, but that came out of Mac OS X, which was itself based on the Mach kernel and UNIX. And Chromebook, likewise, is a stripped-down version of Linux, originally designed to host the Chrome web browser. Point being, simplicity can grow out of complexity, like a flower can grow out of a pile of dung. It just has to be done correctly.

Microsoft’s attempts at simplifying Windows, alas, have not been successful. And it is perhaps instructive to study these attempts to understand why that is so. These missteps explain why Windows is what it is today, and why Microsoft only has Windows whereas Apple has Mac, iPhone, and iPad and Google has ChromeOS and Android.

In Windows Everywhere and the article series that it sprung from, I make the case that the history of Windows as a platform can be seen as a series of reactions. And one of the most visceral of those reactions occurred in 2012, when Microsoft released Windows 8, Windows RT, and Surface.

This was a full-frontal assault on the web and mobile device transformation noted above. It included a new web- and mobile-based app platform called the Windows Runtime (WinRT) and the first full port of Windows itself to Arm. (To be clear, Windows has been running on Arm since the 1990s, but previous versions were subsets of full Windows, products like Windows CE and Windows Mobile.)

In response to the mobile threat—really, the Apple threat seen in the iPhone and then the iPad—Microsoft designed Windows 8/RT to “make your PC work like a device, not a computer,” according to Steven Sinofsky. This was inarguably the right decision, and each of the slightly more specific sub-goals for this release explain how that product team perceived the shortcomings in Windows compared to mobile. Windows 8/RT would be fast and fluid. It would be “touch-first (but not only touch).” It would offer “long battery life,” and “grace and power” in its apps, and those apps would “work together.” And it would be more personal, with live tiles and experiences that “roamed” (synced) between PCs.

That’s a grab bag of good and bad ideas, but we shouldn’t lose sight of how radical this was. Nor should we ignore the incredible engineering effort that went into porting Windows, full Windows, to Arm. Yes, modern Windows has a cross-platform legacy dating back to its inception as NT, but Microsoft hadn’t stretched those muscles in decades, having long ago standardized on the Intel/x64 hardware platform. Bringing Windows to Arm was not straightforward or easy.

As Sinofsky explains in his book Hardcore Software, just calling this effort a “port” is unfair, as that term in no way encapsulates the deep architectural innovations required to bring every Windows subsystem to this completely different architecture. In my own simplistic way, I later noted that the success of this work would be measured by how boring it was. That is, success meant that it just worked, and if it just worked, there would be no drama, no missing bits.

Tied to this, Microsoft had to corral its partner Intel and convince it to add Arm-like innovations to x64 too; modern, device-like power management and efficiency, most obviously. It had to convince PC makers to accept that Arm-based PCs would not be like the x64 platform in key ways that would benefit customers but undermine their ability to differentiate. And it in doing so, Microsoft, would need to control more of the hardware driver stack to ensure a consistent “it just works” experience.

This was never going to be successful.

The first full Windows port to Arm, Windows RT, was undermined by Microsoft’s fear of confusing customers by delivering an iPad-like experience that differed in any way from traditional PCs. By the Microsoft Office team’s refusal to create a mobile app version of its suite on Windows because it already had the desktop suite, and it wanted to focus its new mobile efforts on iPad. By Intel, pushing back again and again and delivering lackluster SoCs like Atom that in no way competed effectively with Arm. By PC makers that wanted more control of the user experience, not less. And by time itself. Three years wasn’t long enough to make all this happen. And so what Microsoft and its partners collectively produced in 2012 was an ugly mess, like every Play-Doh color badly mixed together and extruded thoughtlessly into the world.

To be clear, making the PC “work like a device, not a computer” was the right decision. It’s just that almost all the specifics were wrong. Windows 8 failed. Windows RT failed. The Windows Runtime, later called Universal Windows Platform (UWP), failed. And Surface failed, though it obtained some measure of success three years later with the Surface Pro 3 form factor that continues forward to this day.

The next Windows team tried to clean up this mess, and while I believe it was mostly successful, it’s also fair to say that doing so involved rolling back the radical changes as much as possible and putting the PC back on an evolutionary path that would halt an exodus while not improving its relevance or potential for growth. Windows 10 righted the ship, but it was also a holding pattern. Modern PCs benefit from some of the innovations that happen first on mobile, but the underlying platform remains rooted in the legacy past, and it’s unreliable and inefficient.

Picking up the baton dropped by Sinofsky, the Windows 10 team pushed forward with its own Arm efforts via the newly-renamed Windows on Arm (WOA). Like Windows RT, WOA would target multiple Arm chips, but like Windows RT, it would also consolidate to a single chipmaking partner, in this case Qualcomm, with the idea of expanding to a bigger ecosystem of partners in the future.

WOA would address a key complaint about Windows RT. Thanks to the steady evolution of the Arm platform, it was now possible to emulate x64 code on Arm, a technical impossibility just years earlier. So where Windows RT users were limited to the ported x64 software that came with the device—Windows, its in-box apps, and most of Microsoft Office 2013—WOA users could also download Windows apps from the web. At least in theory: This was limited to 32-bit apps at first, and the emulation always introduced performance issues. Issues that will soon be resolved, we think, by the Snapdragon Elite series chips and a generation of low-level OS evolution.

But WOA doesn’t address an important part of that goal to make Windows “work like a device, not a computer.” WOA, after all, is still Windows. That is, it looks like Windows, quacks like Windows, and works like Windows. It is Windows. Where Google and Apple offer simpler user experiences on mobile, Microsoft still offers the same old complexity. This complexity exists at a low level, because WOA is Windows and must do everything Windows does to run all Windows apps. And this complexity exists in the user experience, with its decades of antiquated cruft.

And that is a very long way to explain the concerns that led me to write this article. Here we are, decades later, and Microsoft and its partners may finally arrive at this future we’ve imagined for so long, a future in which Windows runs well on modern, device-centric hardware. It’s full Windows, real Windows, not some subset of Windows. And the resulting PCs can be fanless and silent, efficient, and reliable. Can work so well, they might in many ways be boring.

And it may not matter.

For all the effort that’s gone into this, for all the time we’ve spent watching and waiting and hoping, this latest effort to push Windows into the 21st century, Windows 11 on Arm running on Snapdragon Elite, may simply represent another holding pattern. It may buy Windows some time, and prevent an exodus, while doing little to change its relative position in personal computing. What it won’t do, can’t do, is trigger a new era of growth. Windows 11 on Arm, WOA, is still Windows. And Windows is still a complex solution for a personal computing market dominated by simpler mobile devices and web services.

I raise this issue in part because we, as a community, have rallied around too many losing efforts over the years. And we could be doing it again in prematurely celebrating the impact of the Snapdragon X Elite on Windows on Arm. Like the owners of the Atlanta Falcons celebrating their coming victory from the sidelines in the closing minutes of Super Bowl LI, I’m worried that we will see our victory stolen away as part of a historic comeback. What if ChromeOS and Chromebook Plus play the role of the New England Patriots in spoiling the victory we assume to be ours?

It’s entirely possible that the simpler, safer, more modern, and lighter weight ChromeOS is a better fit for the Snapdragon Elite hardware platform, and that premium devices utilizing this platform and built to the Chromebook Plus specification might be a more formidable foe for the MacBook Air and Apple’s other Arm-based Macs. It’s worth at least considering.

This thought first occurred to me a month ago when The Walking Cat leaked what I believe was our first look at a Snapdragon X-based PC, the Yoga Slim 7 14(.5) 2024 Snapdragon Edition laptop. It’s a pretty, professional-looking device, thin and clearly fanless and thus silent, much like the MacBook Air. And though you can see Windows 11 on its display, and a Copilot keyboard on the keyboard if you look closely enough, the more I looked at it—and I stared at it quite a bit—the more my brain kept repeating, “Chromebook Plus.”

Weeks later, having seen several other Snapdragon X-based leaks, and having spent lots of time using ChromeOS Flex and a Chromebook Plus device, this still makes sense to me. I do believe, and strongly, that the Snapdragon X is the right chip family for Windows on Arm, and that it’s the start of an overdue hardware correction away from the inefficient Intel processors of the past. It’s a love letter to the fans who swooned for Windows on Arm from the beginning, only to be disappointed again and again.

But the Snapdragon X also solves similar problems for ChromeOS and the Chromebook Plus. Google’s desktop platform is a perfect storm of web and mobile, and it is in many ways better aligned with the efficient, always-connected nature of the Snapdragon X chips. More native, if you will. Chromebooks also align better with the realities of the modern world, where most people spend spend most of their time using phones and only sometimes need a larger display and a full-sized keyboard. Maybe this, combined with on-device and in-browser AI capabilities, is the hardware leap forward that puts these devices over the top.

We’re going to find out. As I’ve noted several times now, I can’t wait to get my hands on a Snapdragon Elite-based PC, and having yet seen any in person, I’m leaning towards a 15-inch Surface Laptop. That could change, but that’s where I’m at right now.

But I’m also wondering now about a 15-inch, thin, fanless, and silent Chromebook running on Snapdragon Elite. How that could pair with my Pixel smartphone. And whether such a setup might be a simpler, more efficient, and more reliable computing experience than Windows—or the Mac—offers. For me, of course, but also for most people.

It’s going to be an interesting year, no matter what happens.