Broadcom

Pays reasonably well (although pay is decreasing).

There are small pockets of good people here and there. Sadly, they are limited in what they can accomplish by the general malaise endemic in Broadcom.

Deadwood - lots of it, and no way to get rid of it.

Can't hire good people because you can't get rid of the bad ones. It doesn't matter anyway because, as a whole, Broadcom never wants to do anything but glob together X different chips in Y new process to keep getting the same low margins for chip Z. You don't need outstanding engineers to do that.

These days, most chips require software (sometimes a lot) to be useful. For the most part, Broadcom is slowly drowning in its inability to create good software for controlling its chips. It is pretty dismal working in software for Broadcom.

No way to advance your career, particularly if you are in a satellite office.

The culture is what you would expect of a 20-year-old company that has grown with little to no vision by randomly acquiring other companies.

Poor cohesion and coordination in all regards.

I realize this is a pretty dismal review, but it depends on perspective. There is certainly a segment of engineers that doesn't mind this sort of culture. Broadcom is a company that has decided that the best way to expand and grow is to acquire companies for new and interesting technologies. There is not even the pretense of trying to do anything but the next increment internally.

The most sad thing about Broadcom is that it is positioned incredibly well to be an innovative powerhouse. It has awesome IP and great fabs. There are also some outstanding engineers. It could become the Google of the chip space. But it won't be.