Hacker News Reader

Comment by doorman2 | Hacker News Reader

Mighty had no hope of succeeding. I’ve built something very similar to Mighty, and the fact is that browsers are hogs, and then adding on video transcoding on top of that makes the situation completely untenable. If someone can’t afford a computer good enough to browse the web, they sure as hell can’t afford a computer that can transcode HD video at all times, which is essentially what they’d need to pay for.

View on HN

This was EXACTLY the problem Mighty was encountering.

They were targeting customers who didn't have powerful enough computers and offering a way to make their computer effectively faster while browsing the web without needing to upgrade their computer.

The problem is the solution costs $35/mo. That's $420 a year.

Anyone who can afford an extra $420 a year to improve their computer speed, probably is just going to buy a better computer with that money.

You can buy a used M1 Air for $700. Maybe $650. It'd be significantly faster than Mighty's solution because M1's ST is faster than Epyc's ST.

Browsers aren't single threaded, so the ST performance isn't the end-all-be-all metric to use. JavaScript is single threaded, but the browser itself uses multiple threads. E.g. there is the main thread, which handles the user input. There is the rendering thread, which actually takes up a majority of the CPU time on a lot of websites. And then there is stuff like GPU acceleration, which makes a lot of the web performant when it wouldn't be otherwise. Of course, a server does all of these things worse than a decent laptop.

It is the end all, be all metric. The reason is because once a certain threshold of MT performance is reached, which the M1 most certainly reaches, then the ST is what matters for how fast your website loads, even with multiple tabs opened.

Reply on news.ycombinator.com