I wish FPGAs and other more purpose built and purpose suited options were available in my IT equipment. They can do amazing shit, better and more efficiently. Just wasn’t ever available to use for me at least.
One problem is that programming an FPGA is a rare skill. A lot of even good programmers simply don’t get their heads around how they tick.
Coming from a digital hardware background, an FPGA was amazingly straightforward, so I’m one of the rare breeds who does both FPGAs and microprocessors.
Tell that to the people who think they will soon replace this expensive and complicated FPGA stuff with something running on a cheap MPU programmed by an intern. For thirty years now…
If you’re at a scale where making a new ASIC is your go-to, congratulations on your job at Google or Apple. I don’t even know if FaceMeta would do that. Designing and founding a new chip is a whole thing.
There are a load of things in IT where using a processor is the wrong choice, and using an FPGA instead would have made a lot of problems a non-issue.
I wish FPGAs and other more purpose built and purpose suited options were available in my IT equipment. They can do amazing shit, better and more efficiently. Just wasn’t ever available to use for me at least.
One problem is that programming an FPGA is a rare skill. A lot of even good programmers simply don’t get their heads around how they tick.
Coming from a digital hardware background, an FPGA was amazingly straightforward, so I’m one of the rare breeds who does both FPGAs and microprocessors.
That’s awesome! I bet that gives you some really good insights to the two platforms.
Is that controversial? I’ve always assumed people avoid FPGAs just because they’re unfamiliar with them.
Tell that to the people who think they will soon replace this expensive and complicated FPGA stuff with something running on a cheap MPU programmed by an intern. For thirty years now…
Yikes.
Welcome to my world.
I think it’s because FPGAs are an intermediate to just making your own ASIC.
If you’re at a scale where making a new ASIC is your go-to, congratulations on your job at Google or Apple. I don’t even know if FaceMeta would do that. Designing and founding a new chip is a whole thing.
Agreed. If you’re not moving hundreds of thousands if units, there probably is no return in investment. Those things have expensive up front costs.