A warning about Xilinx FPGAs is that routing time is quite slow (i.e. can be tens of minutes), though computation is fast once routing is done. They're OK so long as one doesn't need to change the circuit very often...
Hugo de Garis and I played with some neural nets on Xilinx FPGAs in Xiamen, a few years ago.... But then he got convinced it was easier to just use (digital) GPU supercomputers ;) ... ben On Wed, Jul 11, 2012 at 9:01 AM, Steve Richfield <[email protected]>wrote: > Sergio, > > On Tue, Jul 10, 2012 at 1:49 PM, Sergio Pissanetzky < > [email protected]> wrote: > >> you are not alone. How big can one go with an FPGA that is currently >> available? 1K? 10K? 10K would already be nearing some practical >> applications with EI, but 100K would be better. >> > > Take a look at the Xilinx documentation at: > > > http://www.xilinx.com/support/documentation/sw_manuals/xilinx13_4/ug872_largefpga.pdf > > With 28nm fabrication, a million or so analog devices would probably be > about the limit with present methods. > > Depending on the details of their designs, some existing logic circuits > come very close to being able to function as useful analog components, e.g. > I have seen circuits where designers used a simple inverter as a sort of > "poor man's comparator". All sorts of bad design practices emerge when it > become necessary to modify existing circuit boards to fix design bugs, and > do it without adding any additional components. > > I think we may be close to figuring out how to make analog components that > could self-organize and wire themselves, which would provide infinitely > greater capability. > > I am thinking EI because I am sure that, if EI can be demonstrated for >> example in image recognition, then it would attract attention immediately, >> including the chip makers. "General computation" is too vague. Or, better, >> I propose to start "general computation" with EI, then one could expand. >> >> ** ** >> >> Also, personally I believe this would be "hyper-Turing" but I would be >> very careful with that term because there is too much hype about it. Ben >> has strong reasons why it is better not to use the term for now. >> > > Ben or Sergio: Could one of you send me an off-forum message explaining > this? > > I am very happy that such things can be calculated, and there is plenty of >> time to find out if they are hyper or not. **** >> >> ** ** >> >> Do you do these things? Do you build analogs from components? >> > > Sure. I have designed several special-purpose analog computing circuits > for past clients, and as I explained in an earlier posting, I have had > dayslots of hands-on experience using an EAI TR-20 general purpose analog > computer, without which I probably would have never been able to pass my > diffy-Q class. > > I don't have any money, but just saying. >> > > It doesn't take much. Old general purpose analog computers sell on eBay > every month or so for ~$1,000 each. I sort of wish I could spare the money > to buy one. If even that is too expensive, you can purchase a chip that > contains four operational amplifiers for about a dollar, attach some > components salvaged from dead electronic devices, apply power, and watch it > compute. Of course these systems certainly aren't capable of doing anything > the size of what you are thinking, but reading about them and playing with > them certainly gives you an idea what they could do, and how to do it. > > There is a fun game for analog computers - of creating differential > equations to plot intricate fractal-like patterns. I wish I had saved some > of my old drawings. > > Do these fond memories jog anyone else here to remember their long-past > times with analog computers? > > Steve > > *AGI* | Archives <https://www.listbox.com/member/archive/303/=now> > <https://www.listbox.com/member/archive/rss/303/212726-11ac2389> | > Modify<https://www.listbox.com/member/?&;>Your Subscription > <http://www.listbox.com> > -- Ben Goertzel, PhD http://goertzel.org "My humanity is a constant self-overcoming" -- Friedrich Nietzsche ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-c97d2393 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-2484a968 Powered by Listbox: http://www.listbox.com
