Takeaways from SC14

Now that SC14 is “in the books” I would like to share some of my takeaways from a very successful conference.

First of all, the conference is a great opportunity to visit with our customers and talk about their current needs and learn what we can do to help them with their plans for the future. I was gratified by the feedback received regarding our emphasis this past year on quality and testing. We received very positive feedback from customers having recently upgraded to our Torque 5.x and Moab 8.x releases.

Adaptive Computing Booth Demos

During the show I spent most of my time helping out at the Elastic Computing demo in our Adaptive booth. We announced a new “Elastic Computing” feature in the soon to be released Moab 8.1 Torque 5.1 releases which adds the ability to add and delete dynamic nodes to Moab and Torque from an external source to handle “bursty workload.” In our booth we had a live demo where we demonstrated the ability to use the OpenStack interface to spin up VM nodes, add them to Torque and Moab to handle a burst of workload and then return the nodes when the workload demand returned to normal levels. A customer I spoke with from a university mentioned that this is exactly the type of thing that would help them when they do not have the budget to add physical hardware for a project but with Elastic Computing they could include a budget for VMs in their grant proposals and have a much greater chance of getting funding for their projects.

Another takeaway from the conference is the opportunity for serendipitous encounters and discussions where someone shares a challenge they are facing and someone else chimes in with a great suggestion. I was able to attend the Torque BoF and the Torque dinner afterwards and saw this first hand. On another occasion I sat down for lunch in a common area of the conference and struck up a conversation with one of our customers. He asked if our new “Elastic Computing” feature was limited to OpenStack and I explained that since our solution uses triggers and scripts that they can use the feature with any (and even multiple) providers of VMs or even physical hardware. He said that it would fit in perfectly with their plans. He said that this was the 3rd or 4th time at SC’14 where just by chance he struck up a conversation and found a solution for something he needed help with and that this is the reason he loves to come to Supercomputing.

Echoing that same sentiment, the most important things I always take from Supercomputing are the feedback and ideas from customers that will help us meet their needs in the years to come.

What was your takeaway from SC14?

Facebook Twitter Email