The OpenCPI framework was created in its current open source form in 2009, when a U.S. DoD program supported its transition from a proprietary prototype created at Mercury Computer Systems, Inc. (now Mercury Systems, Inc.) to an open source framework licensed under the LGLP3 license from the Free Software Foundation (here). The DoD program, called the Software Producibility Initiative, aimed to improve how software solutions were created in the U.S. DoD, from both a productivity and licensing perspective.
Prior to 2009 and the transition to open source, this technology was created from requirements and concepts developed for real-time processing in broadcast video distribution and real-time processing for medical imaging. It was brought to a usable state after a major effort funded by DARPA in its Polymorphous Computing Architecture (PCA) program in 2002-2007 where it was applied to a new heterogeneous multicore processor architecture called MONARCH, which contained a mix of RISC cores and DSP cores. It was also used in conjunction with the Software Defined Radio mandates in the JTRS (Joint Tactical Radio System) program during 2004-2008, and was experimentally applied to CIED (counter-IED) requirements in 2008-2010.
Since OpenCPI was established in 2009 it has been further supported and enhanced by various simultaneous government funded applied research efforts that continue in 2017.
All these efforts work to achieve the synergy between open source software/gateware, processing technology and vendor independence, and the concepts of component-based development (CBD).