Loading ...
Sorry, an error occurred while loading the content.

Re: the folly of classification, was Re: [midatlanticretro] Vintage vs. just old

Expand Messages
  • Dave McGuire
    ... But the same is true for all compiled languages, even semi-compiled languages like Perl. ... I consider it hardware because, by loading the bitstream,
    Message 1 of 29 , Jul 26, 2012
    • 0 Attachment
      On 07/26/2012 03:19 AM, Dave Wade wrote:
      > On 26/07/2012 07:33, Jeff Jonas wrote:
      >>>> Another example: Is an FPGA-based computer "real hardware"
      >>>> or "an emulation"? If you think it's real hardware,
      >>>> the RAM-based bitstream you squirt into an FPGA
      >>>> is written in a hardware-description language,
      >>>> nearly all of which are Turing-complete programming languages,
      > I think that's a sleight of words (i.e. a bit of magic). Whilst might
      > be VHDL a "Turing Complete" programming language you don't run VHDL on
      > the FPGA you run the OUTPUT of the VHDL on the FPGA and in most FPGA's
      > you can't (well perhaps don't) change the program on the fly.

      But the same is true for all compiled languages, even "semi-compiled"
      languages like Perl.

      > I
      > consider it real hardware because it has the same issues as real
      > hardware. You can get issues with timing, propagation delays, gate sink
      > and source current etc. that you don't get on a software emulation.
      > However as they are not the "same" issues you get on the original
      > hardware in some sense its a clone.

      I consider it hardware because, by loading the bitstream, you make
      permanent (for this power-up cycle) changes to what the hardware does.
      But dynamic reprogramming of FPGAs is happening more and more often,
      making the line (for me) become a bit blurry.

      >> I have 2 replies to that.
      >> One in my normal voice and one in a high squeaky voice :-)
      >> 1) What if the same circuit were made into an ASIC or direct to "real" silicon instead of an FPGA? It would act the same, taste the same. Same output and internal states to the same input. I don't see how initializing the FPGA is part of the process or categorization
      >> 2) Even in the 60s, IBM's logic diagrams were machine generated. Not drawn by hand. I'm unsure of their internal tools, but PERHAPS they were using an FPGA-like language to describe the desired circuits instead of just describing it transistor-and-gate-level?
      >> If that's true, THEN a similar hardware-description language was used to design the ORIGINAL hardware.
      > I am not sure that's true, but the diagrams are incredibly complete.
      > However take a look at this:-
      > http://www.ljw.me.uk/ibm360/vhdl/
      > seeing as he had to add delays I would say that "real hardware".....

      But you can run VHDL and Verilog as "software" on a computer as well.


      Dave McGuire, AK4HZ
      New Kensington, PA
    Your message has been successfully submitted and would be delivered to recipients shortly.