Total Posts: 38
Joined: Dec 2009
|
We looked at Triad's via configurable arrays; do you think field programmable versions on a grand scale could be possible? I'm not talking a few specific things around a standard PLD, I'm talking primarily analog.
Posted on 2010-02-03 14:29:49 at 2010-02-03 14:29:49
|
Total Posts: 14
Joined: Nov 2009
|
Feasible but not necessarily marketable
I think it's possible, especially at lower frequencies. High speed stuff will get extremely difficult with modular architectures. Unfortunately, that's where technology is primarily going these days. Having programmable gain stages, signal routing and filtering all integrated on a single chip seems like a cool idea.
Part of the beauty in using discreet components for analog is the ability to tweak and troubleshoot the circuit relatively easily - given a competent surface mount solderer. Plus analog components are pretty cheap.
Posted on 2010-02-19 17:57:18 at 2010-02-19 17:57:18
|
Total Posts: 21
Joined: Mar 2010
|
I shure hope so
I really hope that one day massive analog arrays are feasible. We do not use high frequencies (20-150kHz), so we are the low-frequency category. Having an analog computation engine would be great, as it is inherent real-time. But the amount of filtering we have to perform is just to large to implement with discrete components. Some groups develop cochleas as VLSI chips, but then you have to make VLSI chips, which is kind of "special" to the non-chip-designers.
Having the same density on an analog array as current FPGA's show would be another "holy grail"
Posted on 2010-03-25 03:25:18 at 2010-03-25 03:25:18
|