Arzeda uses computational models to custom-make enzymes that do not occur in nature. In an effort to affordably scale up the processing power and storage required for some of its complex mathematical computations, the biotechnology company implemented a Windows Azure cloud services solution. As a result, Arzeda has access to cost-efficient, highly scalable, computing and is free to focus on its business instead of its IT infrastructure.
Arzeda is a biotechnology company based in Seattle, Washington, that develops new enzymes not found in nature to produce desired chemical reactions. Its intention is to replace petroleum-based products and processes with more sustainable bioprocesses. The company applies a methodology that integrates the power of chemical catalysis, the high selectivity of biological macromolecules, and the flexibility of computational design.
This methodology requires extensive compute power. Arzeda maintained a small in-house Linux cluster, but found that the cluster did not have enough processing power for some its computational calculations. “Being able to compute large-scale problems is critical to our business,” says Alexandre Zanghellini, Cofounder of Arzeda. “The amount of processing power that we need is on par with what you would find with a 1,000-core cluster.”
Though the company needed a way to scale up its infrastructure to meet its compute needs, its need for more processing power was not constant. Arzeda needs massive scalability only a few days each month, as project cycles dictate. As a result, the cost to build an on-premises infrastructure was exorbitant and unrealistic for the small company. In fact, Arzeda estimated that to build and maintain the IT infrastructure required to handle the computational capacity that it occasionally needed would cost at least U.S.$250,000. “We don’t have a need for static computational capacity,” explains Zanghellini. “So to invest that amount of capital would simply be overpaying for infrastructure that we use only a fraction of the time.” Solution
To address its needs for highly scalable, cost-efficient processing for its computational design methodology, Arzeda decided to implement the Windows Azure
cloud services operating system. Windows Azure serves as the development, service hosting, and service management environment for the Windows Azure platform. It provides developers with on-demand compute and storage services to host, scale, and manage web applications on the Internet through Microsoft data centers.
||We want to focus all of our energy on designing new enzymes—that is our business. With Windows Azure, we can maintain our focus on what we do best and let Microsoft handle the infrastructure.
After enrolling in the Technology Adoption Program for Windows Azure in May 2009, the company worked closely with Microsoft to get its UNIX-based software to run on a Windows-based platform. In September 2009, Arzeda started developing its computing service for Windows Azure and was up and running by December 2009. “Developing for Windows Azure is straightforward, but working with the Microsoft team to port the software from UNIX was a big part of the success of this project,” says Yih-En Andrew Ban, Project Leader and Scientist at Arzeda.
Arzeda uses Windows Azure compute and storage services. Scientists at Arzeda prepare jobs for computation, packaging them in XML messages and sending them to queue storage in Windows Azure. Scientists then submit a job request and start compute instances—anywhere from tens to hundreds of instances—on Windows Azure, which picks up the jobs from the queues and processes them.
The company uses Blob storage in Windows Azure for its data input and data output files, which are also easily scalable. The data input for its jobs is typically in the magnitude of tens of megabytes, but the output is in the magnitude of tens of gigabytes. Arzeda also uses Table storage in Windows Azure to store job-state information.
When the jobs are complete, scientists download the computations to on-premises computers that the company uses for data analysis. Using a sweeper process, Arzeda can automatically shut down its Windows Azure instances as soon as all the jobs are finished processing.Benefits
As a result of implementing Windows Azure, Arzeda has a cost-efficient, scalable solution for its compute-heavy computational transactions, so now it can focus on its business instead of worrying about managing IT infrastructure.
Improved Scalability for Compute-Heavy Processes
When scientists at Arzeda are running compute-heavy processes for designing new enzymes, they can create as many Windows Azure instances as they need to complete the job—with just a few clicks. “We typically use anywhere from tens to hundreds of Windows Azure instances for our jobs,” explains Ban. “However, as we grow and as demand dictates, we have the ability to quickly scale to the magnitude of thousands of instances.”
Avoided Costly Capital Expenditures
Building and maintaining an infrastructure that could handle the company’s computational needs during its peak times would have cost at least $250,000. By using Windows Azure, the company can take advantage of infrastructure at Microsoft data centers and pay only for the resources it uses. “Windows Azure offers tremendous flexibility, and we pay only for what we use,” says Zanghellini. “We’re spending only 13 percent of what would have been our capital expense for an on-premises solution, and have made it just an operational expense. For an early-stage company, when access to capital is very expensive and usually a lengthy process, this is a decisive advantage."
Frees Scientists to Focus on Business
With only four employees, Arzeda does not have extra resources available to maintain a large IT infrastructure. “We want to focus all of our energy on designing new enzymes—that is our business,” says Zanghellini. “With Windows Azure, we can maintain our focus on what we do best and let Microsoft handle the infrastructure. Instead of managing infrastructure, we can explore nearly infinite protein libraries and deliver solutions that are inaccessible with traditional enzyme engineering techniques.”
For more information about other Microsoft customer successes, please visit: