Dr. Alan Gara
BlueGene Chief Architect and IBM Fellow
An IBM Fellow at the T.J. Watson Research Center in Yorktown, NY, Dr. Gara is the Chief Architect for the three generations of IBM’s BlueGene supercomputers. Gara was recently honored with the 2010 IEEE Computer Society Seymour Cray Award for his “innovations in low power, densely packaged supercomputing systems.” Given Gara’s past achievements, we can’t help but wonder what’s in store next for BlueGene from this highly regarded and talented individual.
* * *
Dr. Gara’s insights and approach in designing the BlueGene system embody the spirit of Seymour Cray’s groundbreaking supercomputer designs, which were marvels of integrated, elegant and systemic design. A two-time Gordon Bell award recipient for his scientific work in supercomputing in 1998 and 2006, Gara’s work is also considered largely responsible for the BlueGene supercomputer receiving a national medal of technology and innovation in 2009.
Currently leading IBM’s exascale system research efforts, Dr. Gara is the technical project leader and chief system architect credited with creating the BlueGene systems design. The architecture is based on the need for dense packaging, low power operation, and efficient cooling, with a high mean time to failure (MTBF) to facilitate large hardware configurations. The BlueGene series set new standards for ultra-high performance, occupying the top position on the Top500 list of supercomputers for several years.
Gara not only conceived the low power BlueGene design, he was the driving force behind its realization, which represented a radical reconceptualization of distributed memory parallel systems. Gara identified power consumption and reliability as two of the primary constraints on the continued scaling of supercomputing architecture, something now widely recognized for exascale computing. He then created a design based on low-power SoC nodes, with dense packaging and multiple interconnection networks that scaled beyond anything previously envisioned.
Dr. Gara received his PhD in Theoretical Physics at the University of Wisconsin, Madison before joining IBM Research in 1999, where he has since been leading IBM’s high performance computing architecture and design efforts.