Cloud Computing and Science – Mathematica Embraces Cloud Computing

Scientists are always in need of poweful computing resources to do complex calculations. They have relied on everything from super computer to distributed systems like SETI@HOME. In fact, High Performance Computing (HPC) can accelerate the speed at which science is done and can open up new vistas to scientists. Scientific community has been tapping into HPC by taking advantage of the underutilized computing power in various research institutions, to run complex scientific calculations. National Science Foundation (NSF) has been pumping money into the research on computing infratstructure to develop what are known as Science Gateways (also called as hubs). These hubs offer scientists collaborative websites in the style of Web 2.0 sites and these sites can be tapped to do modelling and other scientific simulations. This saves enormous amounts of time for the scientists and gives them access to powerful computing infrastructure.

In my article about LHC computing, I wrote about how Cloud Computing can be a big enabler for scientific computing. Recently, Wolfram Research, makers of the famous Mathematica software, has announced a new cloud computing initiative.

Wolfram Research announced an initiative today to develop a cloud computing service for users of their flagship technical computing software, Mathematica. This project is a collaborative effort by Wolfram Research, Nimbis Services, Inc., a clearing-house for accessing third-party compute resources and commercial software, and R Systems NA, Inc., a provider of computing resources to the commercial and academic research community.

Nimbis Services will enable Mathematica to run on diverse set of platforms from supercomputers to Amazon Elastic Computing Cloud (EC2). So far, Mathematica has been more of a desktop application with some ability to run on small clusters. Now, the cloud based processing power will be tightly integrated within the existing user interface of Mathematica (with almost no learning curve for the users).

Running Mathematica on Amazon EC2 will be just amazing. Users can run programs and simulations which utilizes parallel processing with tasks on different clusters being independent of each other. Examples of such computations include Protein Folding, DNA Sequencing, Monte-Carlo Simulations, etc.

As I mentioned in my previous post, cloud computing offers tremendous opportunities for the scientific community. By tapping into the virtually unlimited processing power of the clouds, scientists can venture out to solve some of the complex problems they had left on the table due to lack of powerful computing resources. With the advent of cloud computing, they will have very powerful computing resources for cheap. With the current economic downturn and with the bad levels of scientific funding in the past several years, cloud computing can offer some respite to the scientific community. On top of it, scientists can now foray into new territories requiring very high computing resources without worrying about the cost or the technology infrastructure. Mathematica on clouds is a step taken in the right direction and I hope we see more cloud computing adoption in the scientific community.