Mellanox Evening Event at SC11

One of my favorite ‘SC activities has always been the annual Mellanox event. After many years of attending, Mellanox is repaying the favor this year and asked me to be one of the keynote speakers, along with their CEO, Eyal Waldman, and TACC’s Jay Boisseau. I’ve had the opportunity to work closely with Eyal and Jay over the years and they are always interesting speakers.

While I don’t want to give-away too much about my talk, it will definitely be aligned with the evening’s theme of how Mellanox’s connectivity solutions are delivering the key ingredients for sustainable HPC environments. HP will be talking a lot about FDR Infiniband during the SC11 show so you should look forward to hearing from me on how HP and Mellanox are doing delivering FDR solutions to help customers drive down the cost/byte and power/byte of HPC networking.

For many years, the cost and power drivers in HPC where on the compute side, and while $/watt/FLOP are still dominant, technologies like Nvidia’s Nvidia’s Tesla GPUs and Intel’s upcoming Many Integrated Core technology delivery drastically improved performance/watt over traditional x86 processors. As this trend continues, as we reach Exascale performance targets by the end of the decade, FLOPS essentially become free and the power and cost to move data outside of the CPU starts to dominate.

Already today, FDR and many QDR networks use optical cables to reduce the power required to move data between racks of servers. Somewhere around the middle of the decade, it is expected that interconnects within a server will turn to photonics to drive down the power and cost of moving data within a single server. By the end of the decade, exascale systems are likely to have photonic interconnects built directly onto the processor chips.

To hear more about my thoughts on Infiniband and HPC, be sure to register today for the Mellanox event at SC11. Hope to see you there.

About these ads

About Marc Hamilton

Marc Hamilton – Vice President, Solutions Architecture and Engineering, NVIDIA. At NVIDIA, the Visual Computing Company, Marc leads the worldwide Solutions Architecture and Engineering team, responsible for working with NVIDIA’s customers and partners to deliver the world’s best end to end solutions for professional visualization and design, high performance computing, and big data analytics. Prior to NVIDIA, Marc worked in the Hyperscale Business Unit within HP’s Enterprise Group where he led the HPC team for the Americas region. Marc spent 16 years at Sun Microsystems in HPC and other sales and marketing executive management roles. Marc also worked at TRW developing HPC applications for the US aerospace and defense industry. He has published a number of technical articles and is the author of the book, “Software Development, Building Reliable Systems”. Marc holds a BS degree in Math and Computer Science from UCLA, an MS degree in Electrical Engineering from USC, and is a graduate of the UCLA Executive Management program.
This entry was posted in Uncategorized. Bookmark the permalink.

One Response to Mellanox Evening Event at SC11

  1. Hi Mark,

    I’m intrigued by your post – “Exascale performance targets by the end of the decade, FLOPS essentially become free and the power and cost to move data outside of the CPU starts to dominate.”

    Is you presentation from the Mellanox event available?

    Regards,

    John Mitchell
    Active Cable Solutions Marketing Manager
    Intersil Corp.

Comments are closed.