Within our Symbiotic research programme we are developing two embedded processors. Polymer is a parallel multiprocessor for computationally intensive applications, and Monomer is a simple parallel uniprocessor for applications where space or power is very limited. Despite their small size, both designs are code compatible, capable of around 150MHz operation on mid-range FPGAs, and offer incredibly easy yet sophisticated high-performance parallel processing operation.
The processors are bristling with architectural features aimed at achieving small circuit size, high speed, configurability, programmability, high code density, and efficient use of resources.
Both processors are configurable at synthesis time to support almost any data width and addressing range according to the application. To further save circuitry and power consumption, if an application rarely uses certain operations (e.g. multiplication or long shifts) then these can be easily moved out of silicon or replaced by slower and more compact circuits.
A parallel computing architecture, scalable from embedded systems to supercomputers, efficiently handling fine-grain concurrency levels of hundreds of thousands.
Variants include 16- to 64-bit general purpose systems, transmuteable instructions, and arbitrary precision arithmetic.
A case for a new architecture