> Just did it on my 2011 MacBook Air. Pushes all four cores up... about 15%.

15% of 4 cores ~ 60% single core capacity. That's distinct from saturating a core both qualitatively and quantitatively, which means the prediction is not a bullseye... but it's on the order of magnitude in impact, which means they're not entirely wrong.

True, but it's an eight-year-old computer. And it wasn't exactly a computing powerhouse when it was released.

Try this, http://www.voots.name/

That one actually did better.

Two cores went up 20%. The other two didn't budge. Go figure.

There we go. Try that with video formats.

Giphy manages to consume a much larger fraction of my modern desktop's CPU. I think it may have something to do with the small size and multiple copies of the gifs on the page you chose.

For best results, use a Slack channel, or a forum reply page with an array of animated smileys.

To complain about GIF usage saturating your CPU when running an Electron app is to be penny-wise and pound-foolish.

Slack doesn't exactly have the most efficiently implemented front-end interface.

Ah, yes, I didn't realize that Slack's platform reimplemented the whole graphics-formats-rendering stack. I naively thought that Electron and Slack inherited that from Chromium, but guess stupid me is way behind the times in this technicky stuff. Of course it's all in Javascript now, what was I thinking!

Do things change if each gif is different?

Try again using a gallery page on a site like giphy which is going to be more representative of the use case op was talking about.

edit: such as https://giphy.com/sports/2019-stanley-cup-playoffs which seems to be entirely gifs if you view page resources in the debugger.

I don't believe these are actually gifs here. Similar to sites like imgur or twitter they convert it to something else to host for other people. Probably to keep bandwidth down. You can get to the actual gif, but you have to dig for it.

Most animated images on giphy are vp8 videos.