Modular computation is a new form of digital arithmetic that will change the way we think about numeric calculation. It’s based on several important disciplines in mathematics and computer science, including modular arithmetic and residue numbers. However, unlike these previous fields of study, modular computation allows for fractional arithmetic much like a floating-point processor. The result is a new field of computer science for which many discoveries still lie ahead.
Many advantages exist with modular computation, including very high accuracy, efficiency and resiliency. These advantages combine to form unprecedented capabilities not contemplated in the prior fields of study. MaiTRIX has been hard at work exploiting these new properties and discovering amazing results!
There are many unique properties of modular computation which differ significantly from fixed-radix arithmetic, such as binary or decimal arithmetic. One significant difference is there is no carry from digit to digit. This property results in unforeseen advantages including radical new circuit topologies which partition circuitry in a much more advantageous manner for certain types of problems. As a result, modular computation provides advantages for 3D IC technology as well as quantum computing. Moreover, modular computation can be implemented now using off the shelf FPGA or standard ASIC technology to provide unprecedented performance for massive product summations integral to applications such as Artificial Intelligence, Digital Signal Processing, Matrix Arithmetic, Cryptography and Financial applications.