A systemic computer is a true parallel computing machine. A living organism is a great example of a systemic computer. A conventional computer, designed according to the von Neumann architecture, is a very, very poor example of a systemic computer. Its centralised ALU and memory, its synchronised and sequential processing, and its brittle, fault-intolerant behaviour are all exact opposites of the systemic computer. While many attempts are being made to parallelise these conventional architectures with multiple cores, there is no man-made computer architecture today that can behave like a real systemic computer.

This causes many problems when we wish to use a systemic computer - basically we cannot make one using today's technology. So to date research has focussed on exploiting the technology of today in order to simulate a systemic computer.

Initial work investigating systemic computation created virtual machines that run on conventional computers. These enable the properties of systemic computation to be investigated, but they run slowly because a sequential computer is being used to simulate parallel systemic computation. More recent work has focussed on using FPGAs or GPUs to optimise the speed of our virtual systemic computers by parallelising some of the computation. A list of different systemic computation implementations follows:

  • 2005 Simple systemic computation virtual machine and compiler, by Bentley.
  • 2006-2010 MS Windows based virtual machine, compiler, run-time environment and visualisation engine, by Le Martlelot.
  • 2009-2010 Cuda GPU optimised virtual machine based on Bentley's model, by Rouhipour and Shayani.
  • 2009-2013 Virtex-6 LX240T FPGA optimised virtual machine, plus simulator and compilers, by Sakellariou.
  • 2013-2014 OpenCL multicore optimised virtual machine, updated language, compiler and visualiser, by Bentley and Deng.

See the downloads section of the wiki for resources, and the publications section for detailed technical information.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License