INRIA and KALRAY strengthen their scientific collaboration on the MPPA MANYCORE processor and parallel computing

Orsay, Paris, 25 March 2013 KALRAY SA and Inria strengthen their scientific collaboration on KALRAY’s MPPA MANYCORE™ processor architecture with the launch of 9 new projects. These will be spread across different Inria research centers and cover a panoply of scientific fields, ranging from purely IT-related subjects (compilation, palatalization, etc…) to the implementation of advanced algorithms in genomic or cryptographic applications.

Created in 2008, KALRAY is a company that designs and markets an entirely new generation of massively parallel processors, simultaneously giving high performance and low consumption for the professional electronics industry (image, audio, and signal processing, communications infrastructures, and industrial monitoring) as well as scientific computing applications and digital simulation.

Joël Monnier, CEO of KALRAY, emphasises the importance for KALRAY of this collaboration with Inria’s teams: "Our MPPA MANYCORE processors are based on breakthrough technologies, some of which originally came from Inria, such as our floating point computing unit for instance. The scientific skills of the nine Inria teams involved in this new strategic partnership will, firstly, allow us to progress on subjects directly linked to programming and the architecture of the MPPA MANYCORE processor, and secondly, thanks to our 256 core processor , offer new computing methods for the scientific and industrial fields of the future, such as genomics, cryptography, simulation, etc."

François Sillion, Director of the Grenoble Rhône-Alpes Inria research centre, reminds us that "the support of innovative SME’s through the exchange of research results is one of the missions of Inria. Furthermore, the multiplication of the number of cores in processors is a revolution with profound and long-lasting repercussions in micro-electronics. It naturally drives our institute to rise to the challenges that are offered by massively parallel architectures. KALRAY’s MPPA MANYCORE processor, being one of the most advanced in the world in this domain, constitutes for us a source of inspiration on diverse subjects of research and experimentation. Our mission is to be at the forefront of these new massively parallel computing techniques, which allow not only science to move forward but also applications for use in industrial and consumer applications that are ever more demanding in terms of computing power".

The subjects covered as part of this national collaboration and the different Inria teams involved are the following:

  • Programming models for new parallel architectures and parallelisation of automatic or assisted coding:
    • AOSTE, Inria Sophia-Antipolis Méditerranée
    • CAMUS, Inria Nancy Grand Est
    • PARKAS, Inria Paris Rocquencourt
  • Very-high-speed, low consumption data encryption:
    • CARAMEL, Inria Nancy Grand Est
  • Run-time environment and distributed processing:
    • COMPSYS, Inria Grenoble Rhône-Alpes
    • DOLPHIN , Inria Lille Nord Europe
    • DART, Inria Lille Nord Europe
  • Data Mining
    • CEPAGE, INRIA Bordeaux Sud-Ouest
  • Supercomputing applied to the understanding of the genome
    • GENSCALE, Inria Rennes Bretagne Atlantique

"The work carried out as part of this national agreement will yield results for direct application in KALRAY hardware and software solutions as early as 2013. Long-term strategic collaborations are currently under way between KALRAY and Inria . We are already envisaging new collaborations that will swell the ranks of the community using our MPPA hardware and software solution" concluded Laurent Julliard, KALRAY’s manager in charge of research partnership relations.

ABOUT KALRAY

KALRAY is a fabless semiconductor & software company developing and selling a new generation of manycore processors for high performance applications. Our products are optimum for low to medium volume series and boost the development of innovative digital applications in fields such as image and signal processing, telecommunication, industrial automation, data security & networking, or transport.

×
Semiconductor IP