Life Sciences “Gets It” -- Pulls Grid into Mainstream IT

Life sciences companies have long held a special appreciation for the advantages that powerful, leading-edge technology solutions provide. After all, the larger pharmaceutical companies are investigating hundreds of different diseases at any given time and sifting through databases of millions of promising compounds and proteins that might lead to a cure, or at least a treatment. Universities and medical research organizations are equally engaged.

Research is their life blood, and the cures or treatments they discover not only save lives but also determine each company’s viability and financial success. So these companies incessantly search for practical ways to leverage computing assets and speed innovation.

At United Devices, we serve a broad spectrum of industries including pharmaceutical, telecommunications, manufacturing, energy, government and academia. However, we entered the market with an exclusive focus on pharmaceutical companies – which lasted for years. As a result, United Devices is now the undisputed leader in providing computational platforms in this sector as evidenced by the fact that seven of the ten largest pharmaceutical companies in the world rely on our solutions. Besides enabling these companies in a commercial fashion, we have also pioneered discovery work in global research grids where we have played a key role in landmark studies relating to cancer, smallpox, anthrax and the human proteome folding project, among others. In fact, United Devices owns and operates the world’s largest philanthropic grid (www.grid.org) which contains over 3 million devices that span over 200 countries.

United Devices has learned much from these associations. Our people work daily with the best and the brightest within leading life sciences companies. As a result, United Devices has a unique and privileged view of where the life sciences industry is going related to information technology.

Trends in Life Sciences IT

HPC strategies vary across life science companies, however, there are overarching themes. Whether it involves grids, virtual clusters, HPC resources or tapping into the enterprise data center, all roads lead to simplified infrastructures, rapid deployment of systems when additional capacity is needed, improved flexibility in matching resources to applications, compelling reductions in total cost of ownership, and improved productivity.

Besides reducing cost, the value proposition most often cited by executives as the primary driver for HPC strategy is application acceleration – the need to run jobs in minutes or hours instead of days or weeks. This capability “reenergizes the drug discovery process,” they say.

These overarching themes are consistent across HPC and mainstream IT environments – specifically in the data center and managed services arenas. In mainstream IT, however, the ability to automatically provision resources based on application requirements and policy definitions takes priority and application acceleration takes a back seat.

Life sciences companies (including clinical research organizations, pharmaceutical corporations and biotechs) face a number of serious information processing challenges, including ever higher volume and density of data, growing complexity of IT ecosystems, and competitive and regulatory pressures. With the increased libraries of data available for processing in areas like drug screening and PK/PD analysis, researchers must make tough decisions about reducing scope of work to accommodate project timelines. This comes at a time when these scientists have to deal with a growing complexity of IT systems and tools and ever increasing regulatory mandates.

United Devices Solutions: Traditional R&D

United Devices has helped address these challenges in life sciences by offering a range of solutions designed to meet the most pressing business challenges related to HPC infrastructure management. These solutions include:

• Multi-cluster Meta-Scheduler for workload management of geographically dispersed clusters and overflow computing resources.
• Virtual Cluster for transparently extending existing dedicated cluster capacity.
• PC Grid for harvesting underutilized cycles from non-dedicated resources.
• Cluster Scheduler for workload scheduling and management on dedicated clusters
• HPC Analytics for reviewing workload and system performance profiles across heterogeneous clusters to support better decision making.

Of course, each of our life sciences customers is different, with unique needs driven by the size of their operations, the nature of the projects they are working on, their mix of technology assets, organizational and geographic factors, and even cultural preferences. So they assemble and customize these solutions in different ways to satisfy their unique requirements.

The Move to Grid in Mainstream IT

Most of the top tier pharmaceutical companies now have large production grids in place that consist of thousands of devices. With grid technology risk mitigated and the business value proven, these large companies are driving United Devices to leverage its grid core competency to provide solutions in mainstream IT. The most notable requirements are application virtualization, and automated provisioning and management of virtual and physical (bare metal) machines.

United Devices formally announced our Data Center One and Service One solutions last year. We had been co-developing these virtual infrastructure management solutions in concert with key industry partners for more than two years, and our pharmaceutical customers were in the vanguard of the effort. These new capabilities are strategic and are being well received.

Today, few people are surprised that life science companies are among the first to leverage the competitive advantages that grid and virtualization technologies offer. Their research organizations are providing leadership in actively funding grid initiatives on a large scale.

Resource-oriented initiatives like consolidation, standardization, and server virtualization have proven to reduce capital costs in the data center. However, in most cases, the labor and operational costs have risen. For example, higher level skills are required to manage the complexities associated with the proliferation of virtual servers. Intelligent automation is the only way to control these operational costs and ultimately advance to a lights-out data center scenario. From a technology perspective, grid is a compelling way to automate the provisioning and management of the underlying data center assets. The technology is hardened, proven to scale to thousands of devices, and has a natural services orientation.

Based on our experience working with customers in the data center world, we designed our solution to be application-centric – not resource-oriented. Policies reflecting business priorities, service levels, and user and application access to assets are defined and maintained, and resource provisioning actions are automatically executed based on policy-evaluated events.

Such solutions must work across the full range of hardware, operating systems and networks that a customer might be using. United Devices has always worked hard to develop truly heterogeneous solutions that map to legacy requirements, and our virtual infrastructure management solutions follow that same approach.

Data Center One

United Devices’ Data Center One solution is an application-centric infrastructure virtualization solution. For the first time, enterprise data centers can automate the process of provisioning business-critical applications across shared pools of physical and virtual assets. Data Center One uses our core grid technology, a new policy engine, and a new repurposing engine to dynamically allocate physical and virtual IT resources according to SLA-mandated performance targets and other application specific requirements. No other virtualization vendor can offer this level of automation today.

The result is that, with Data Center One, IT shops can significantly increase resource utilization rates, remove operational bottlenecks, further consolidate resources, reduce infrastructure and application TCO, and enable SLA-driven performance. No where are these improvements more apparent than in SAP implementations. In fact, United Devices and Satyam issued a white paper that profiles the 35% infrastructure cost reductions that can be achieved when using Data Center One to automate a typical SAP environment. And we have virtualized other enterprise tools like SAS for major pharmaceutical companies.

Besides all the benefits for the enterprise data center, Data Center One also is able to dynamically unleash selected enterprise IT assets for use by research organizations that were historically fully dependent on “siloed” computing resources.

Service One

Service One uses the same core technologies as Data Center One to enable IT service providers to deliver business applications using a Software-as-a-Service (SaaS) delivery model. The solution enables automated, SLA-driven provisioning of applications to multiple customers by utilizing shared pools of IT assets that may span data centers, business units and companies with no restrictions on location or ownership of assets.

Global telecommunications companies are driving the early implementations of Service One because grid is network aware. This enables them to leverage their connectivity footprints as a key differentiator when competing with standard IT service providers. However, Service One also is garnering attention within large IT organizations in pharmaceutical companies that want to offer business applications as a service to multiple business or operating units. And adoption by contract research organizations (CRO’s) is sure to ultimately benefit the smaller life sciences companies as well.

Summary

Working with its life sciences, biotech and pharmaceutical partners, United Devices developed and hardened its core grid technology in the HPC environment. Other industries followed and embraced solutions from United Devices to optimize and manage their own HPC infrastructures. United Devices now offers the same type of scalable, yet practical solutions for mainstream IT, and life sciences is once again leading the charge.

Obviously, demand for dedicated HPC environments will continue to grow, and innovations from United Devices will continue to lead in the areas of global enterprise grids, virtual clusters and dedicated computing. Furthermore, United Devices will continue to power the world’s largest philanthropic grids such as www.grid.org and IBM’s World Community Grid (which is built on United Devices technology).

At the same time, a number of mainstream IT use cases have emerged in data centers and managed services environments that require the same fundamental capabilities inherent in grid technology that have proven so fruitful to researchers in life sciences. With direction and help from its pharmaceutical customers, United Devices is now providing data center automation solutions for mainstream IT.

IT professionals around the globe are watching what happens in the life sciences industry because they know that when it comes to advanced computing technologies, life sciences “gets it.” And, as always, United Devices will be right at their sides to turn vision into practical reality for life sciences, and for all the industries we serve.