You are here: Home / Science / RTLWS Submitted Papers / 
2020-01-29 - 08:24
Details of the Real Time Linux Foundation Working Group Project

OSADL Project: Real Time Linux Workshops

Real Time Linux Foundation Workshops since 1999

Real Time Linux Workshops

1999 - 2000 - 2001 - 2002 - 2003 - 2004 - 2005 - 2006 - 2007 - 2008 - 2009 - 2010 - 2011 - 2012 - 2013 - 2014 - 2015

15th Real Time Linux Workshop, October 28 to 31, 2013 at the Dipartimento Tecnologie Innovative, Scuola Universitaria Professionale della Svizzera Italiana in Lugano-Manno, Switzerland

Announcement - Call for participation (ASCII)Hotels - Directions - AgendaPaper Abstracts - Presentations - Registration - Abstract SubmissionSponsors - Gallery

Controlled thermonuclear fusion and hard real time Linux: the International Tokamak Experimental Reactor control architecture

Simone Mannori, ENEA C.R. Brasimone

Linux “-soft” and “–hard” real time variants are extensively used for the supervision, data acquisition and real time control of the International Tokamak Experimental Reactor (ITER, www.iter.org). The ITER control room is equipped with Red Hat Linux workstations running EPICS (Experimental Physics and Industrial Control Systems) as SCADA. A large part of the data acquisition and control systems black-boxes used for thermonuclear plasma stabilization are embedded PCs running Linux kernels patched for hard real time operations. We describe the general architecture of ITER and the details of the typical hard real time black-box as formalized by CODAC (Control, Data Access and Communication) team. Finally, we explain the still open issue of the intrinsic incompatibility between closed source “binary blob” drivers (required for GPU computing applications) and the hard real time Linux patches.

Despite the recent “middle finger” initiative by Linus Torvalds, both NVIDIA and AMD still distribute the Linux driver for GPU computing applications as binary drivers not compatible with the available Linux hard real time patches. This incompatibility represent a disturbing “show stopper” for the people involved in plasma real time control algorithm development (that require computing power in the Teraflops range) and close the door to other very promising hard real time HPC (High Performance Computing) applications like dynamic turbulence reduction and on-line cavitation prevention.