what is unary computing?
Unary computing is a new, unconventional paradigm in digital computing in which an “uncompressed” data representation (unary) is used instead of the traditional binary representation to make computations very efficient, with limited hardware resources to perform complex calculations. Conventional methods would require many multiplications and addition operations to calculate functions such as exp(x), tanh(x) and y=x^g through polynomial expansions. Unary computing is able to perform these calculations cheaply and directly without any multiplication operations or additions. These functions are widely used in applications in signal processing (image / video / audio / radar signal processing). Furthermore, unary computing can be used in significantly reducing the computation complexity of constant coefficient multiplication, which is the core of many machine learning (ML) applications. As validated by our experiments, reducing computation complexity increases throughput (how fast computations are done), reduces hardware area (size and cost of the chips to perform computations), and reduces power (indirectly affects power consumption costs, and cooling costs). These are important metrics that have direct impacts on industry costs.
The proposed method is compatible with existing design methodologies and tools: all the measured costs include conversion from binary to unary, cost of computations, and cost of converting unary results back to binary. As a result, given the right tools, designers can integrate our computation methodology in existing digital designs because they can treat our computational units as black boxes that take binary inputs, and deliver binary outputs.