By Scott Bradley and Jon Milby
The College of Arts & Sciences is making strides in its representation of computational sciences, complementing recent faculty recruitment efforts in several departments with a new computing environment designed to meet the needs of researchers.
The scale of available computing systems has often limited computational researchers. Supercomputing environments such as those owned by UK and other national organizations have impressive resources available, but are not always a practical option for some types of research. These systems are designed to run continuously and at capacity, creating queues that may make it impractical to run smaller workloads or test new algorithms.
At the other end of the scale, individual workstations address availability issues, but are inherently limited in the amount of resources they can provide. A well-equipped workstation often lacks the computing power required to run even moderately complex simulations in a practical timeframe. The new computing cluster in A&S will strike a balance between these two extremes, providing researchers with access to expanded computing resources while retaining a great degree of flexibility and availability.
The new College of Arts & Sciences Research Computing cluster (which made its campus debut in the summer of 2014) shares the basic design elements of a modern supercomputer, though at a smaller scale and lower cost. Multiple systems are linked together within a high bandwidth, low latency framework, allowing researchers to run demanding applications across hundreds of processors simultaneously.
The cluster is approximately one-tenth of the scale of a typical supercomputer, but offers resources several times greater than what can be offered by a single workstation. This new system will provide researchers with access to large-scale resources with little to no wait time. By addressing this need and removing common obstacles to research computation, A&S faculty will be able to perform their research more efficiently.
Q&A with Mark Meier, Department of Chemistry
Q: How has your field of research evolved over time to get to this point?
A: Ten years ago, it was inconceivable that you could computationally handle interactions between proteins and solvent, ions and water, modeling 20,000 water molecules at the quantum level. But now, the code and the computational power is there to allow us to look at the question: Can we model the real world in that kind of detail on systems that are reasonably large?
Q: How exciting is it to see the new computational hires coming together?
A: It will be very interesting to get our new computational hires in Chemistry, Biology, and others together to talk. These people have tremendous tools to bear. It’s very exciting to put them all in a building together, close to the Department of Biology and the College of Pharmacy, allowing us to begin to get interactions back and forth between people that see the problem in the clinic, and the people who can model the problems. There’s potential for some tremendous synergy with the location of the Academic Science Building being at the interface between Chemistry, Biology, Pharmacy and the clinical units in the medical center. The same goes for interactions between Chemistry and the Center for Applied Energy Research.
Q: What’s the main bottleneck right now as far as the amount of computational research that can be done?
A: People. UK Chemistry did not hire computational people during a generation when other research universities did, so we’re a bit behind in this area. I’m really excited about Pete Kekenes-Huskey and Chad Risko - they are perfect fits for UK. They will strengthen the Chemistry Department along with a number of other programs.
Q: How do you expect the Academic Science Building to affect Computation Research?
A: This is going to have a tremendous, positive influence on a number of aspects of the research program in Chemistry and around the university. We’re going to be able to gather together people like the computational folks and their collaborators within Chemistry, to get some feedback and to discover new areas for collaboration.
A beautiful facility like this will have an enormous impact on our ability to recruit the best and brightest graduate students, the best and brightest future faculty, and the best and brightest undergraduates. My firm belief is that the ASB will have a very positive influence on many aspects of the research enterprise at UK in addition to the obvious benefits for the teaching program, for decades and decades to come.