Managing IT Infrastructure for HPC in 2011

To start off the new year, I thought I would make some predictions about how organizations will manage their HPC infrastructure in 2011.
In many organizations HPC infrastructure has historically been managed outside of the core IT group. Be it high-end Linux clusters or proprietary systems, HPC infrastructure was often considered too complex and non-mainstream to be handled by an organizations primary IT group and was managed by specialized HPC admin staff. However, as Linux becomes increasingly mainstream in IT departments, it is now more common for HPC infrastructure to be managed by core IT staff. In fact, for many organizations, HPC increasingly is their core IT function. Mainstream HPC management tools such as HP’s Cluster Management Utility (CMU) continue to make it easier for IT staff to manage HPC clusters without having to delve into unsupported open source tools and utilities. So be it DIY with dedicated HPC admin staff, or part of the traditional IT department, in-house management of HPC infrastructure will certainly continue in 2011 as an option for many organizations.
Cloud was for sure one of the most over-used terms of 2010, but in 2011 we will surely see more cloud-based HPC offerings and more organizations turning to cloud based infrastructure for HPC. When Amazon released their first HPC Cluster Compute Instance offering for AWS in 2010, many hard-core HPC users didn’t see it as a serious HPC offering. But Amazon is definitely serious about HPC. They have followed up their first HPC offering with a Cluster GPU Instance and in fact posted a substantial
231st on November’s Top500 list. It is safe to predict that HPC cycles run on AWS and other cloud providers will increase in 2011.
Managed HPC Services
While HP is perhaps better known for helping customers like
Tokyo Tech build PetaFlop supercomputers, HP is increasingly getting into the business of managing IT Infrastructure for HPC. Two large IT management deals announced in December by HP where the BP win and as well as a multi-billion dollar award at NASA. While these deals are not HPC specific, they are likely to touch large parts of both BP’s and NASA’s HPC infrastructure in 2011 and beyond. Even if no new IT management awards were announced in 2011, helping customers manage their HPC infrastructure, and cycling that experience back into the hands of HP engineers working on new HPC products, is set to increase in 2011.

So OK, maybe no groundbreaking predictions for people to challenge, but no matter if you plan to move HPC to the cloud, continue to manage your HPC infrastructure yourself, or look to managed IT services for HPC in 2011, HP is there for you.


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.