Related

Timbre has been used throughout the years on many projects as a tool. Without it, there would have been less projects and less work done. In particular, a related line of work that has been underway for many years is the creation of a hardware platform that will eventually be Timbre. This started with a simple stack computer and progressed as better hardware tools became available. While the end is not yet here, the finish line is getting closer. These projects have sometimes taken a life of their own as course projects and some of them are useful in their own write. Several documents from these projects are available for downloads hopefully to inspire the reader. The DFP has been posted as a website and is also available through opencores.org. Licensing information is available on the website and it has been used in commercial products. The Granular Computer Architecture is the latest incarnation which moves beyond any fixed structures and looks toward vertical integration of hardware and software minimizing resources to only those required as well as opening the door to distributed granular computers.

  • A Stack Computer - A graduate class project report on original work on writable computers using a stack computer as an architecture and VHDL as an assembler. Partial instruction sets for Java and PostScript are also presented. Simulation was done using QuickVHDL tools. Topologies are also discussed.
  • A Stack Computer: Synthesis and Appendices - A graduate class project report on secondary work on writable computers changing to a pseudo-harvard architecture where the instruction register has a feedback loop. Statistical addition is well covered and used as a test program. The latter half of the report provides a review of several Synopsis digital synthesis and simulation tools.
  • A Writeable Computer - This paper is a short summary of both the above projects and was presented at the 1997 Rochester Forth Conference.
  • Neuroprocessor Modeling - This paper details how a neuroprocessor network can be built from stack processors
  • Neuroprocessor - using a stack computer to implement two implentations of a neural network
  • 8-bit stack Processor with Perl compiler - a detailed look at the writable 8 bit processor implemented in the above paper with an assembler written in Perl:
  • Data Flow Processor (DFP) - an evolution of the stack computer design with no inherent instructions
  • Granular Computer Architecture (GCA) - an IEEE paper showing the merging of software and hardware methodologies for implementing real time computational intelligence algorithms.