Green Flash : energy efficient real-time computing for ELTs

Speaker : Damien Gratadour

Abstract : Controlling sophisticated AO systems on the future ELTs is a challenge which is yet to be resolved. As the ELTs are adaptive telescopes, this is critical to all ELT instrumentation. Green Flash is an international EU funded joint industrial and academic project intended to study and exploit future and emerging computing technologies for ELT scale AO real-time control. This includes the hard real-time data pipeline, the soft real-time supervisor module as well as a real-time capable ELT-scale simulation to test and verify proposed solutions. To date, we have initiated in-depth studies on GPUs, MICs and FPGAs. Moreover, beyond computing capabilities, it is critical for these future AO RTC to address the data flow challenges both in terms of the large rate of streaming data from sensors and in terms of heterogeneous data streams in the system. A strong emphasis is thus made in the project around interconnect strategies using dedicated hardware, middleware and software. With this R&D program we aim at feeding the E-ELT AO systems preliminary design studies, led by the selected first-light instruments consortia, with technological validations supporting the designs of their RTC modules. The culmination of the project will be an on-sky demonstration of a chosen solution. However, all proposed solutions will include a detailed study of capabilities and scalability. Here we review the project goals and the results from the studies at the mid-point of this three year endeavor and we describe the downselection process that will lead to the design of a full featured prototype to be eventually implemented in the lab and tested with the real-time simulator.