Cloud computing has gotten a lot of attention in the last couple of years. As a result, countless companies (such as Amazon, Rackspace, etc) now offer services for those who need quick access to servers and storage- without the need for server rooms, full time sys admins, etc.
So with all of these pre-existing services for doing cloud computing, why are we creating our own? There are a few reasons, but one important reason is that scientists often work with BIG data. When a user needs to upload and/or download hundreds of gigabytes, or even terabytes of data from the cloud, they are limited by the network speeds between their computer and the cloud. Within the university, we have very good network speeds, which makes it feasible to transfer large datasets to and from servers. What may take hours to transfer to the Amazon Cloud may take minutes to transfer to the Genome Center Cloud.
Service and expertise are another place we hope to add value to the cloud experience. The Bioinformatics Core has a lot of experience installing and using various scientific software, which we can use to save users hours of frustration when trying to get their research going. Other services are great, but their only aim is to provide computing resources. The Bioinformatics Core is here to help you do science!
At the moment, most cloud computing services are aimed at more “run of the mill” users. Bioinformaticians are a demanding bunch, and when they need a high memory (think 500GB or RAM) or powerful (imagine 50 cores) machine to do an analysis, commercial solutions fall well short (generally they don’t offer more than 70 GB of RAM or 8 cores).
We are currently working on building the necessary infrastructure, and are aiming to make the cloud available to customers by fall 2013. If you are interested in using the Genome Center Cloud, please email us at firstname.lastname@example.org.