The twelve-year overnight success

In October 2012, Microsoft launched the Surface RT: a thin, light tablet running Windows on an ARM chip, the same kind of power-sipping processor found in phones. There was one problem. It couldn't run the Windows programs people already owned.

Nine months later, Microsoft took a $900 million charge on tablets nobody wanted.

By any normal standard, that's where the story ends. Idea tried, idea failed, lesson learned.

Except Microsoft kept trying. A second generation arrived some years later. It was slow and buggy, and the less said about app compatibility the better. Then, in 2024, a third generation built on Qualcomm's Snapdragon X chips showed up with all-day battery life, real performance, and most applications simply working.

Twelve years. Three tries.

What changed

The idea didn't. It was right in 2012: a laptop that lasts all day and stays cool. What changed was everything around it.

• The chips caught up. ARM processors now sip power like a polite guest at a tea party and still handle real multitasking.
• The software caught up. More applications are built for ARM, and the rest are translated far better than before.
• Someone proved it could be done. When Apple moved the Mac to its own ARM chips in 2020, buyers and manufacturers saw that it didn't have to be a compromise.
• The industry committed. Major PC makers that once sat it out now sell ARM models.

Why it matters in the field

For anyone who works away from a desk, this is a practical change. Think of inspectors, field crews, and people in buildings where finding an outlet is a scavenger hunt. A device that lasts the whole shift, stays cool, and can be patched and secured remotely is a real improvement.

Too early, or just wrong?

This is the hardest call in technology. When an innovation fails, it may be a bad idea. It may also be a good idea that arrived early.

Some never get there. I've written about voice-first computing, which keeps arriving and leaving because the obstacle is human behavior, and better chips won't fix that.

A few questions help tell the difference:

• Was the idea wrong, or the execution? Missing apps can be fixed. People's reluctance to talk to a gadget on the bus is harder.
• Is anyone still investing after the failure?
• Has someone proved it works somewhere?
• What exactly would have to change, and is it changing?

What to do with this

Don't write off a technology because version one flopped. And don't bet the budget on version one either.

Put a few devices in the hands of real field staff. Test the applications you can't live without, because compatibility still isn't universal. Expand if it works.

If you haven't looked at ARM laptops lately, you might be missing a platform shift. Plenty of good ideas need a third try. That's what iteration looks like.

An earlier version of this article first appeared on LinkedIn. Join the discussion there.