“Starting Up” with Cloud Computing
Cloud computing has long since left the “buzzword stage” of its life cycle, with most major companies integrating it into their technology and workflow to some degree. Cloud computing is often described through a number of analogies — from a “data lake” capable of storing all the data your firm has collected, to a cluster of virtual computers constantly popping in and out of existence to meet your needs.
The Cloud Computing Service
Data centers — housing phenomenal numbers of servers — are the core of cloud computing. These servers are either made available to clients directly (known as Infrastructure-as-a-Service, or IaaS), or divided into virtual units pre-loaded with the necessary software for your application (known as Platform-as-a-Service, or PaaS). Many times these servers are specialized for certain tasks; some are capable of storing large volumes of data, others have high-performance processors for demanding workloads, and still others have specialized hardware, such as graphics cards for rendering and GPU computing.
So what is the advantage of using ‘cloud’ servers, and what makes them unique? The centralization of these servers in a single data center reduces the burden of infrastructure costs by taking advantage of economies of scale. These machines can then be rented out for only the time and capacity that the particular client requires. As a company using cloud services, this means that you are only paying for the storage and processing power you need at the moment, leaving you flexible to scale without wasting capital.
Cloud Computing Utilization
An excellent example of a company that leverages cloud resources to its advantage is Slack, the company responsible for the messaging application of the same name, which is used by workplaces across every industry. Soon after its launch, Slack experienced a dramatic surge in usage. Traditional scaling would require not only predicting future computing and data storage needs, but also enormous spending to build the infrastructure to meet those needs in-house. In order to scale rapidly, they instead turned to Amazon Web Services’s public cloud. By utilizing AWS’s compute instances, storage buckets, and relational databases, they had immediate access to a robust infrastructure capable of handling their growth.
The business world is embracing cloud computing, and the sciences are hopping on board. The availability of faster processing and massive storage volume is especially well-suited to scientists’ needs. In many fields — including bioinformatics — experiments previously only dreamed of are now possible. The PredictProtein suite, developed at the Rost Lab in 1992, was once a useful, if limited, sequence analysis and function prediction web service. The near-limitless storage and elastic processing of cloud computing, however, have allowed the service to expand its database and fluidly adapt to demand. Now, it can analyze a query sequence with exceptional speed, returning a wealth of information about the protein — including its secondary structure, solvent accessibility predictions, nuclear localization signals, disulfide bonds, unstructured loops, residue flexibility, interaction sites, family hits, homologous sequences, and more.
Cloud Computing at Macromoltek
Here at Macromoltek, computing requirements vary dramatically at different stages in the project pipeline. At times, we might need a large number of concurrent processors and threads to handle parallel workflows. In other situations, we rely on fast, powerful processors to run our simulations. We take advantage of a variety of different hardware options, but establishing a traditional on-premise system to cover every possibility would be cost-prohibitive and extremely inefficient. Not only would there be the need for expensive infrastructure — and the technical expertise to maintain it — but those assets would also be seriously underutilized when they are not in focus. Mixing cloud services with traditional systems gives us the flexibility to rapidly scale up our capabilities to respond to new challenges.
In addition to our production pipeline, cloud resources are used extensively in Macromoltek’s development process. The plethora of hardware and performance options available in cloud instances lets us test our algorithms in a variety of virtual environments, so we can make informed purchasing decisions for on-premise systems. Furthermore, computationally demanding programs can be moved to the cloud to avoid interrupting the development workflow.
Cloud Computing is the New Standard
Using cloud resources has become a common strategy for growth-oriented firms and brings exciting new opportunities within reach. The possible future applications of cloud-centered scientific and enterprise computation are boundless. Cloud computing provides scaling flexibility and an enhanced capacity to take advantage of new technologies. This makes cloud computing an invaluable tool for any company and the standard for the future.
Links and Citations
1. https://aws.amazon.com/solutions/case-studies/slack/
2. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3533974/
Looking for more information about Macromoltek, Inc? Visit our website at www.Macromoltek.com
Interested in molecular simulations, biological art, or learning more about molecules? Subscribe to our Twitter and Instagram!
