Since the beginning of the multicore era, parallel processing has become prevalent across the board. However, in order to continue a performance increase according to Moore’s Law, a next step needs to be taken: away from common multicores towards innovative many-core architectures. Such systems, equipped with a significant higher amount of cores per chip than multicores, pose challenges in both hardware and software design. On the hardware side, complex on-chip networks, scratchpads, hybrid memory cubes, non-volatile memory and stacked memory, as well as deep cache-hierarchies and novel cache-coherence strategies will enrich the current research areas in the future.
However, the ROME workshop focuses on the software side because without complying system software, runtime and operating system support, all these new hardware facilities cannot be exploited. Hence, the new challenges in hardware/software co-design are to step beyond traditional approaches and to wage new programming models and operating system designs in order to exploit the theoretically available performance of future hardware as effectively and power-aware as possible.
The focus of the ROME workshop stands in the tradition of a successful series of events originally hosted by the Many-core Applications Research Community (MARC). Prior MARC Symposia took place at ONERA research center in Toulouse, at the Hasso Plattner Institute in Potsdam and at the RWTH Aachen University. Starting in 2013, the organizers continued this series by establishing ROME as one of the co-located workshops of the Euro-Par as the prime European conference for parallel and distributed computing.
While the 1st ROME workshop, which was hosted at the Euro-Par 2013 in Aachen, was still a MARC-related follow-up event but for a broader audience, the 2nd ROME workshop, held in conjunction with the Euro-Par 2014 in Porto, already expanded its focus to research questions arising from a dawning many-core dominated exascale era.
In 2015, this broader focus was essentially retained for the 3rd ROME workshop, which was held in conjunction with Euro-Par 2015 in Vienna, but the relevance of runtime and operating system aspects was stressed once again as being the primary scope of the ROME workshop series.
This year, too, authors from all related disciplines are invited to submit unpublished papers regarding software for novel many-core hardware architectures. The call for papers especially emphasizes on the challenges and research questions arising from the upcoming generation of heterogeneous and/or massive parallel systems stepping towards a many-core dominated exascale era. The topics of interest include, but are not limited to:
The workshop is scheduled on Tuesday, August 23, 2016, in room 201 as half-day workshop.
Session 1
Short break
Session 2
Coffee Break
Session 3
Workshop papers must not exceed ten twelve single-spaced, single-column pages (LNCS style). On acceptance of the submission, at least one author is required to register for workshop attendance at Euro-Par 2016 and present the paper in the workshop session.
Upload your submission to our submission server in PDF format. The link to the server will be published soon. It must not be simultaneously submitted to the main conference or any other publication outlet.
For the workshop, we will prepare hand-outs with the accepted papers. The revised versions will be published after the conference in the workshop proceedings of Euro-Par 2016, part of the LNCS series of Springer.