Ask Paul: April 12 (Premium)

Happy Friday! Let’s kick off the weekend a bit early with a great set of reader questions. Not surprisingly, many of you are wondering about the same topics I am.
Don’t worry, be happy
madthinus asks:

I want to take you to task, oh dear, an Angry Dear Paul letter!

Gulp.

Twice now you have mentioned that Google has shipped an official Chrome for Windows on Arm because of the coming chipsets from Qualcomm and Twice now I have rolled my eyes.

What is more likely: Google porting Chrome to native ARM because Windows on Arm is going to take off, or Microsoft did most of the optimization work for Edge on Windows and Google are piggyback off that?

That sounds plausible, and I certainly don’t have any insider information about how or why Google uncharacteristically chose to support this platform. But more on that in a moment. I do have one pertinent piece of data that may be relevant to this discussion. Or not.

In the briefing last week, Qualcomm discussed the performance advantages of native Arm apps over their x64 equivalents on Intel. For web browsers, the measure was the Speedometer 2.1 benchmark, which is no longer the most current version, but still interesting. On Arm, Edge performs fully 57 percent faster than it does on Intel, which is a big different. But Chrome and Brave saw smaller (but still meaningful) improvements of 20 and 15 percent, respectively. And perhaps because I’m a Brave fan, I asked about that. Why were Chrome and Brave so similar to each other, but so different from Edge?

I was told that the Edge team was obviously invested in this shift, and that Brave, in this person’s view, was just a “skin over Chrome” (or similar, I forget the exact phrase). He never mentioned Chrome explicitly, perhaps because Qualcomm and Google had partnered on the announcement. But the inference was that Chrome and Brave saw similar gains because they are architecturally similar to each other, whereas Brave (and thus Chrome) is not as similar to Edge.

Does this suggest or prove that Google thus did not just piggyback on Microsoft’s work? Not necessarily: If you think back to the news about Microsoft bringing ClearType text rendering to Chromium, and thus to Chrome, you can see that Chrome uses a different text renderer than Edge. And it’s equally possible that differences like that explain the performance gain differences between Edge and Chrome/Brave on Arm. (I assume Brave uses the Skia text renderer like Google.)

Also, you don't think Qualcomm showed their PC chip to Google for Chromebook support? And Google saw the value in making optimizations in Chrome and ChromeOS for a view on a next generation Chromebook?

This is also plausible. And, as noted above, these two companies did partner on the announcement: The original PR outreach was very clear about that, and the date changed a few times because of changes in Google’s schedule. Chrome was “fully optimized” for Snapdragon-powered PCs, i...