We recently had a biotech company contact us about setting up a dedicated system to be hosted at our data center. Although this particular request was for a relatively small, private infrastructure, often research companies need to conduct sophisticated calculations that can be best achieved through cloud computing. We will first look at the biotech connection to the cloud, then specifically discuss a dedicated real world scenario.
Introduction – biotech and an “Aha” moment
With the growing dependency of the biotechnology and pharmaceutical industries on cutting-edge cloud computing strategies, one can wager that we are nearing a future in which highly sophisticated, personally customized drugs will be available within days. Genomic testing has become much more advanced through the rapidfire performance and availability of the cloud. When oncologists in search of a cancer cure and specialists throughout medical science have run data (for diagnostics, efficacy of treatment protocols, etc.), they have previously been largely dependent on how much processing and computing resources can be allocated for their projects. Now, power for any data algorithm is immediately accessible, along with costs that are much more affordable.
It should be noted that excitement about cloud computing is not merely the realm of cloud hosting providers and software-as-a-service companies: many researchers, data scientists, and business professionals from all segments have started to have an “Aha” moment, in which they are starting to realize the full potential of cloud computing. A March 2013 report by the Association of American Medical Colleges (AAMC) argued that cloud computing should be taken very seriously by the biotech industry and biomedical scientists. Reporter Stephen G. Pelletier noted that it’s understandable how incredibly IT systems have grown within medical research, but running a massive datacenter internally may not always be the best choice, and the legacy model can in some cases be prohibitively slow and expensive. The cloud just makes sense.
In layman’s terms, just how fast is the cloud? The AAMC interviewed Geoffrey C. Fox, PhD, an associate dean of Indiana University’s informatics and computing college.
Dr. Fox’s point of view is particularly noteworthy because Indiana University is the home of Big Red II, a supercomputer capable of handling one petaflop maximum, equivalent to one quintillion floating point operations each second – 1 million billion calculations every single second. That rate matches IBM Roadrunner, the first supercomputer to achieve a petaflop in 2008. As of November 2012, per Ars Technica, the speediest supercomputer in the world was the 17.6-petflop-capable Titan, located at the Oak Ridge National Laboratory in Oak Ridge, Tennessee.
Dr. Fox told the AAMC that biotech was a perfect match for cloud computing since it didn’t require the particular configurations of supercomputers. The cloud is basically as elastic as you want it to be: you can have hundreds or thousands of servers process your request at once, on-demand. The availability of the cloud, though, is particularly compelling: “The cloud’s spare capacity often enables it to process a researcher’s data faster than a supercomputer, “said Fox, allowing researchers to minimize wait times.
Real World Scenario – custom dedicated biotech system
We work with medical research teams and biotechnology professionals every day. Here is an interaction that occurred between one of our consultants and an anonymous biotech client:
Client: We would like to purchase HPC servers according to our configuration (super micro with 1-2 TB RAM) and host it remotely with SSH access. I am looking for quotes for this project to be done in the next month or so. Thanks!
Consultant: Thank you for contacting Atlantic.Net. Please provide us with the following information so we can supply you with a formal proposal.
- Number of Cores in the processor required
- Actual Amount of RAM required
- Total Amount of Storage and type of storage ( SATA / SAS / SSD ).
Client: Please find the detailed specs for servers we would like to buy and host: [omitted].
Consultant: Thank you for the information. We have attached the pricing for the computing nodes based on a 12- or 24-month agreement. Before we proceed, we would like for you to review the pricing we have prepared and provide us with your feedback.
The promise of the biomedical cloud
Johns Hopkins announced in January 2014 its involvement in a project to compile a massive digital library of pediatric MRI scans. The “Google-like” portal will allow doctors to more accurately diagnose children suffering from brain diseases, thereby improving treatments.
That’s just one example of how recent cloud technologies are a game-changer for the biotech field. We are excited to be a functional part of this movement to find rapidfire and potentially customized disease treatment methods and cures. For that, we have our un-managed Cloud systems, where you can build whatever data environment you want. However, if you need a custom system as the above firm did, we specialize in dedicated HIPAA Compliant Hosting Solutions.
By Moazzam Adnan