This week in Rhode Island, Dell is sponsoring the 26th Annual High Performance Computing Conference. It’s an exciting week for us, as we know that access to computing power is crucial to companies and organizations looking to innovate, solve complex challenges and stay competitive.
Just such a business – RES Group, one of the world’s leading renewable energy project developers – uses Dell’s high-performance computing solutions to locate sites on which to build wind farms, turbines and solar plants. The company, featured in our new Efficient IT campaign, is on a mission to build a greener future through the development of renewable energy sources, and technology plays a huge role in that.
Q. From your perspective, how is HPC accelerating innovation in the renewable energy industry? About what are you most excited?
A. Wind energy project economics rely upon a good knowledge of the “wind resource,” i.e., how windy it will be on a wind farm over the long term. Traditionally, the wind energy industry has relied upon costly field measurements to assess the wind resource of a proposed wind farm. In contrast, HPC lets us create virtual observations using state-of-the-art numerical weather models. Drawing upon evermore HPC power allows our weather models to offer a more complete representation of the underlying physics. This means we can be confident that our virtual observations are consistent with real measurements. Virtual observations are much lower in cost than real observations and can be obtained far faster. Before we finance and build a wind farm, we will always make real observations at some stage, but virtual observations allow us to delay this investment until we are sure that the proposed project is feasible.
Q. Data center efficiency is imperative to all organizations, but given your field, its impact on energy consumption must be paramount to RES. In your experience, what HPC technologies or strategies are driving the most energy efficiency today?
A. To run our current models, we needed to ramp up our HPC resource by a factor of 20. A simple scaling of our old technology would have had prohibitive electricity usage and cooling requirements. RES is committed to minimizing its own environmental impact as evidenced by our ground-breaking low carbon headquarters. Unless we could find a more efficient technology, our plan to expand our HPC resource would have been deadlocked. Working with Dell, we identified an alternative in the form of Dell PowerEdge M610 blade servers. This gave us the increase in computational resource we needed, but with 75 percent less energy consumption than the simple scaling up our old technology. The impact on running costs of using the more power efficient technology is also very substantial.
Q. In this case study, you say your HPC cluster’s efficient management solution has given your staff more time for strategic tasks. What projects have you been able to tackle with the resources you’ve freed up?
A. We are constantly looking at ways to optimize our workflow so we can spend more time on our most value-add activities, such as weather modelling, and minimize the time spent on other things (such as managing the cluster). Using the cluster management software has made running our new larger cluster simpler than running our old smaller cluster. This is a great result, as when we first looked at our expanding our HPC resource, we were concerned that it could generate a lot of work in terms of management. Thankfully, this hasn’t been the case and we have been left free to concentrate on our weather models.