Many startups end up being acquired by the more mature semi companies. They would have warrants in the startups giving them some control and if the startup were to be acquired by someone else a share in the profits.
"Fantasy" is the better word choice, not "concept", Chris.
Semiconductor startups are competing for VC money with snot-nosed kids with a website. The kids have no chance of, and have no idea how to, generating any revenue, but they are "capital lite" so they get money.
Meanwhile you show a very credible, conservative, 25:1 ROI actually building something, a great management team, and it's "see ya later" because you need $40M to do it. and then there's the bar set by the likes of Tabula, where few VCs believes you can do a semiconductor startup for under $100M
You hit the nail on the head @ElPresidente...since you need $100M to make something happen, you need $1B of TAM for your product and there are very few opps that would give you that TAM (and they are all taken by Broadcom, Qualcomm, and others)...the era of silicon start-ups is finished!
but dinosaurs are not dyeing, Qualcomm, Broadcom, etc are doing just fine, thank you, and growing bigger...it is newer animals that want to rival dinosaurs that are dying out...mammals can nibble on smaller projects (IP, piece of the design outsourced from the dinosaurs, EDA tool, a testbench) but are not able to take on a large IC breakthrough project
Blog Doing Math in FPGAs Tom Burke 23 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...