https://arstechnica.com/gadgets/...
I have to be honest. I like AMD, but this is really a big fuck up in my opinion. Just helped a guy debug this issue and find he has this problem.

I also really like AMD, but I wouldn't buy new stuff from them before it has proven itself in the field. I'm still on a Phenom-2 1100T because the Bulldozer right after that was expectedly crap.

I was interested in the 2400G APU, but they botched it up again - it was never really stable under Linux. I don't like Linux that much, but I won't buy hardware that would block this route from the get-go.

This year, I'm eying the 3400G which seems to work with Ubuntu 19, so I'll wait for Mint 20.

Sloppy. AMD still seems a little too willing to put headline grabbing stats over quality. Didn't touch ryzen or epyc because of the bulldozer fiasco.

With any luck they'll get comfortable at some point and start focusing more on quality and intersections of value and what users actually need.

So AMD has fixed this quickly and the author isn't sure if their board is asus or asrock? Sounds like an article written by an ars.

> October 29, 2019
that one they fixed in the linux kernel in less than 3 days?

also lots of yall buy intel's paper-plate-security shit the instant it comes out but yall gonna wait on that AMD shit for one bug that was almost instantly fixed???

@Parzi There's a little bit more software out there than the Linux kernel.

@electrineer @Parzi I was planning on building an AMD system do stuff like this happen a lot and if so are they quick to fix it?

@Bubbles it's rare for AMD but every other month for Intel. Usually it's fixed with software (somehow???) pretty fast and then BIOS updates fix it at a firmware level within a year.

But how does this get missed before production? Unit tests?