A Startup’s View of HPC For the Masses

If there is a new CPU, GPU, storage, or networking technology coming out in the next 18 months, chances are HP has already designed it into a future HPC product. We know that when Intel talks about their new 3-D transistors being used in future Ivy Bridge processors, there will be customer demand for HPC servers based on that technology. Every now and then, however, I like to take a slightly different look into the future and there is no better way of doing that than taking a look at where venture capital companies are putting their money.

Last night I had the intellectual pleasure of attending General Catalyst Partner’s Entrepreneurs Forum, along with about 700 others representing startups and those interested in investing in startups. Ray Ozzie kicked off the evening with a short talk about the booming opportunities being brought on by the rapidly evolving growth of mobile devices and cloud computing, followed by Diane Greene talking about her startup days at VMware.

I talked to several dozen companies, ranging in startup lifecycle from two grad students with an idea to approaching $1B in sales. One thing was universal, everyone was working on either mobile or cloud based products. There wasn’t a single startup I talked to that wasn’t creating a product born in the cloud. If you are a startup and want funding, you better be able to explain how your idea will help build the cloud, run on the cloud, or access the cloud.

One young entrepreneur, also an MIT PhD student, asked me what made me excited about working on HPC at HP. She listened intently as I talked about not only building some of the world’s fastest supercomputers, but doing so using HPC technology for the masses, starting with a single GPU enabled HP ProLiant SL390s and growing it into TSUBAME2, one of the fastest, and more importantly the most efficient, production supercomputer in the planet. She then thought for a moment, and asked, “aren’t you worried Amazon EC2 will just take all your business away from you in the future? I hope more people run HPC in the cloud, I answered, it is a lot easier to sell to 10 HPC cloud providers buying 10,000 servers each than to sell to 1000 end-users buying 100 HPC servers each.

HPC in the Cloud is perhaps the ultimate expression of HPC for the masses. If even a few of the startups I talked to last night are successful, I’m sure we will see a steady stream of HPC cloud advances in the coming years. And that is great news for HP. Not only is HP a leading systems provider to most of the top public clouds, but we continue to invest in leading edge, purpose-built HPC servers that continue to push the limits of performance and efficiency with some of the world’s fastest supercomputers like Tokyo Tech’s TSUBMAME2.

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.