Modern computing devices have processors with multiple processing cores, called multicore processors. Curiously, programming concepts used for multicore programming today were invented already decades ago, in a time when processors had only one core.
These unsophisticated concepts make multicore programming complex and error-prone. As a result, the majority of software programmed today still uses only one core, leaving the remaining cores structurally underutilized. Such underutilization becomes increasingly problematic in the future (as numbers of cores will only increase), decelerating the technological progress that modern society has become accustomed to. Simplifying multicore programming therefore constitutes a major challenge for today's software industry, providing exciting research opportunities for computer scientists. In this Rubicon project, I aim to develop an innovative programming language to simplify multicore programming, built on a pivotal &emdash; recently discovered &emdash; design principle: familiarity-centricity.
Supported by: NWO Rubicon
Participating CS members: Sung-Shik Jongmans