HPC in the Cloud: Of Clouds and Mass Spectrometers: The Varian Case

Dedicated to Covering Enterprise & Scientific Large Scale Cloud Computing

HPC in the Cloud >> Features

June 18, 2010

Of Clouds and Mass Spectrometers: The Varian Case

Nicole Hemsoth, Managing Editor, HPC in the Cloud

The march into the cloud on the part of traditional HPC areas has been slow and gradual overall with tentative adoption here and there, mostly on a private cloud or very selective hybrid basis. One area where there is some significant traction, however, seems to be the life sciences sector—not just for genomic or drug research, but also for producers of life sciences equipment. After all, security and privacy are not necessarily major concerns but the need for on-demand provisioning of vital compute resources is.

Palo Alto-based Varian Medical Systems, a large manufacturer of medical devices and software focus on cancer treatments and medical solutions related to radiotherapy, radiosurgery, proton therapy is known for its informatics software for the management of cancer clinics and their practices. They are also suppliers of medical equipment used in scientific, industrial and care settings, not to mention a supplier of imaging products. HPC applications are diverse throughout the company and include the need to employ simulation to create precise designs for fine-tuned medical equipment.

Varian is one firm that has burst into the cloud on a selective basis based on time-critical demands for one of its products. In one case, they needed to be able to simulate a design for a mass spectrometer, which is a very compute-intensive operation that would require several thousand compute hours and would take an estimated six weeks to run on internal processors. In order to meet a deadline for deliverable results, however, this timeframe was not feasible, thus Varian had to look outside of its normal operations to find a way to increase computer power to speed along the process.

Although there are other cloud management providers who offer similar services, Varian settled on Cycle Computing to handle the task of making sure the mission-critical project could be realized without the significant delay caused by its in-house IT environment. Cycle provisioned a secured cluster on EC2 that they scaled up to execute the simulation before shutting down when the calculations were complete. In EC2, running such a calculation on one machine for 100 hours costs the same as running 100 machines for one hour, which meant that in this case it was possible for the researchers to use and pay for only what they needed before returning to normal operations. This is beauty of the cloud model for short-term needs like Varian’s in this case and is the same model that has been employed across other industry areas during times of peak requirements. According to Cycle Computing, this required 18 hours to become fully functional—and that was with a few unexpected application-centered hitches that required refinement.

In a phone interview with HPC in the Cloud, Jason Stowe, CEO of Cycle Computing noted that, “In the case of Varian, it was a matter of scale. We were able to complete this in 18 hours. We spun up in approximately 400-500 cores and ran the calculation in 18 hours in comparison to the internal cluster they were running on. We were able to get an expontential performance increase, which was largely based on the pleasantly parallel kinds of computation they were doing.”

Stowe’s point is important here because the Varian case exemplifies two ways that their needs are aligned with pushing into the cloud with their mission-critical computation. First of all, this was pleasantly parallel computation, thus the complexities that are involved in several others that are in the strict HPC domain were not barriers. Secondly, security was not as much of an issue as it might have been if they were handling the personal information as some life sciences companies do on a daily basis. All of that can be considered critical to business operations and given the web of compliance complexity involved, it might not make sense to utilize a service such as that offered by Cycle, Platform, Univa or others who provide cloud management solutions.

According to Stowe, there is great movement in the life sciences industry as it is one of the key sectors where outsourcing large amounts of data processing that is not business-critical is acceptable. Since so much of the work of biopharma companies in particular do involves general calculations versus the processing and storage of personal data, it makes sense to push some jobs into a public cloud like EC2.

This is an interesting business model that might have an impact on overall time-to-market for companies both within and outside of Varian’s industry. In order to meet a mission-critical goal, IT work with a vendor to provision specialized clusters that rely on Amazon’s EC2 with the added benefit that it will come standard with application sets that are in the range of the life sciences industry (proteomics, computational chemistry and bioinformatics). The concept behind this offering is to spin up a cluster and make it so the cluster environments change in size according to load and demand, which means that they do essentially the same thing a public cloud would do but with greater control and industry specificity. According to Cycle, this service “offers analysis pipelines that incorporate software favorites such as Gromacs, Bowtie, HMMR and Blast.”

While cloud has been a hot topic of debate in the life sciences community for some time, finding major research firms and institutions willing to climb atop the idea in a dramatic way can be a bit of a challenge—especially since the commercially-oriented outfits have little to gain by revealing their secret tech sauce. Clearly there is some major movement in this industry to look at tentative cloud adoption as there have been a number of veteran and rookie companies clamoring to be heard in this space. Life sciences research is at the cutting edge of technology and the market alike, at least according to IDC estimates of the fastest growing sectors in HPC, so it comes as no surprise that there is some significant noise coming from tech companies with solutions specific to the life sciences industry.

The problem that life sciences-oriented cloud management firms seek to solve is not necessarily in helping them make a full transition to EC2 or another public cloud. Since there are still hurdles that few life sciences organizations are willing to leap at this point (and yes, we’re talking privacy, security and compliance here) cloud management becomes critical when these firms have on-the-spot compute-heavy issues to hammer out but lack the time and performance boost their existing clusters are capable of kicking out. Since EC2 seems to be the top choice for life sciences companies, cloud management operations allow smooth transition into and out of the public cloud to meet occasional bursts in capacity needs by providing the application pre-configured, security-focused ability to do so. In other words, the cloud is perfect for non-business-critical but enormous tasks.

Article Tools

Share Options

(Digg, Technorati, more)

Subscribe

Subscribe to HPC in the Cloud

Discussion

There are 0 discussion items posted.

RSS Feeds

Feeds by Topic

Feeds by Content Type

Around the Web

Application Flexibility and the New Enterprise Architecture

Aug 31 | Application delivery strategies must be shaped with flexibility in mind as the number of platforms delivering core applications is bound to change with time. Since a greater number of devices and platforms are entering the infrastructure mix, those who do not adapt quickly face being locked into strategies that do not mesh well with new developments. Read more...

A Maturation Stage for Storage Virtualization

Aug 27 | Storage virtualization has been gaining momentum as it moves from concept to practice but some suggest the offerings in this realm have not matured sufficiently and require a longer maturation process before wider adoption occurs. Read more...

HP Seeks to Boost Cloud Service Automation Capability

Aug 27 | Although it was lost in the chaos of the 3PAR bidding war between, HP announced news that it acquired cloud service automation firm Stratavia to bolster its cloud management offering and further its strategy in the arena. Read more...

NASA CIO Weighs in on Nebula, Public Clouds

Aug 26 | In an interview from the NASA IT Summit last week, the agency's CIO, Linda Cureton weighs in on developments with Nebula platform and the adoption of the open source code by other agencies looking to the cloud. Read more...

Facing the Realities of Private Cloud Building

Aug 24 | While private clouds are getting far more attention than they received at the beginning of the cloud buzz boom, the realities of the complexities of actual building them--not to mention the financial and time investments--are often overlooked. Read more...

Read more headlines...

Featured Whitepapers

Increase IT Performance from the Enterprise to the Cloud with WAN Optimization

Aug 30 | | Enterprises face a paradox today: while workers become increasingly distributed, IT infrastructure is rapidly consolidating. Virtualization has made it possible to create consolidated, elastic pools.

Grid Computing in the Cloud White Paper

May 14 | | Empower business users, scientists and researchers with their own grid computing infrastructure in the cloud.

View the White Paper Library

Multimedia

Webcast: Storm Clouds - Security Issues in the Cloud and How to Address Them

This Webinar will highlight the four critical areas of concern when securing cloud infrastructure services and managed enterprise applications.

Webcast: Confronting HPC Cloud Computing Security Concerns

Escalating energy and operational costs of building and maintaining data centers are forcing enterprises to adopt cloud computing models. But are Infrastructure as a Service (IaaS) solutions like IBM's Computing on Demand (CoD) really cost effective? Join the discussion as industry experts discuss how you can exploit cloud computing for maximum ROI.