At the J. Craig Venter Genomes, Medicine and the Environment Conference here, SGI today unveiled the SGI® Scientific Workflow Solution. Comprised of some of the world's most powerful, scalable and flexible technologies, the new solution is aimed at simplifying the complex task of conducting breakthrough scientific research by accommodating the various ways in which scientists work.
The new offering integrates SGI's acclaimed clusters, supercomputers, Field Programmable Gate Array (FPGA) products, data management tools and storage arrays into a unified, workflow-friendly, and cost-effective solution.
The SGI Scientific Workflow Solution addresses an increasingly costly problem in today's research labs. To enable new breakthroughs in biosciences and other disciplines, modern scientific workflows must incorporate an ever- widening array of applications -- something that puts extraordinary pressures on organizations struggling to make do with a single primary computing resource.
"Scientists need to focus on their research, not on determining how to modify their gene sequencing routines so they will run on a cluster with 68 instances of Linux," said Michael Brown, sciences segment manager at SGI. "Rather than force-fitting all applications and workflows onto a single computing architecture, scientists are far better served by a high-performance computing solution that provides a uniform end-user view, but can be easily customized to the needs of each user and field of study. The SGI Scientific Workflow Solution delivers that kind of flexibility by leveraging technologies that have evolved from two decades of outfitting the world's most demanding scientific users."
The SGI Scientific Workflow Solution combines award-winning SGI® hardware and software offerings in a standards-based environment that provides an exceptional level of flexibility, enabling facilities to run multiple applications on the same central resource comprised of integrated and optimized computing platforms. With the new workflow solution for scientific environments, both desktop and server systems share a single pool of data, while a common user and system management reduces personnel expense.
The new solution is built on a unique combination of enabling technology elements, including:
* A common Linux® operating system across all platforms, so users need
learn only one environment;
* Platform-wide storage solutions to simplify data sharing, access,
movement, and management across systems;
* A single system to schedule jobs and manage workflow across all systems;
* A centralized systems management interface and operational dashboard for
updating, provisioning, configuring, and monitoring; and
* Optional system specific performance and tuning enhancements through
SGI® ProPack® for Linux.
The new solution is underpinned by an array of award-winning SGI products that together address the vast spectrum of scientific workflow needs. For instance, the new value-priced SGI® Altix® XE cluster line, based on dual- core Intel® Xeon® 5000 Series processors, is ideal for high-throughput and screening workflows. Dual-core SGI® Altix® 450 and 4500 systems scale to thousands of processors and hundreds of Terabytes of data in a single system image, making them ideal for deriving new insights from highly scalable and floating-point or memory-intensive applications, and those that require extraordinarily fast I/O.
The SGI solution also incorporates the groundbreaking SGI® RASC(TM) (Reconfigurable Application-Specific Computing) technology, which enables application-specific co-processors like FPGAs to tap into an Altix system's shared memory address space at full system bandwidth. SGI RASC blades can be reconfigured on site to drive optimal performance from specific algorithms -- such as gene sequencing, image processing, image segmentation and computationally intensive data analysis -- without having to scale the number of CPUs.
"The scientific community has stringent requirements for high-performance computing systems that are optimized under a number of different platforms and architectures," said Anders Dellson, CEO, Mitrionics, Inc. "SGI's Scientific Workflow Solution provides customers with systems specifically designed for varying functional needs, performance, and pricing. SGI's recent addition of FPGA Supercomputing technology to its portfolio gives the company an even stronger value proposition and one that is unique in the industry. Mitrionics has been very pleased with its close partnership with SGI to develop and deliver FPGA-accelerated technologies and applications to the market."
SGI storage, archiving and data management solutions also ease the process of giving all users access to a single pool of scientific data. The wide range of SGI® InfiniteStorage arrays are complemented by solutions for centralizing critical data and streamlining data management, no-compromise data sharing over networks without replication, and transparent transfers of rarely used data to near-line and offline storage alternatives.
"The BioTeam believes that it is critical that workflow be friendly to scientists and yet provide an extensible informatics platform to express and publish operational procedures to face everyday productivity challenges," said Michael Athanas, BioTeam co-founder and scalable computing specialist behind iNquiry, a ready-to-use cluster enabled web portal optimized for use in life sciences. "SGI is addressing these challenges with the common access and management environment available in the SGI Scientific Workflow Solution."
The SGI Scientific Workflow Solution will be offered in several versions optimized for major fields of study, including bioscience and earth science. SGI will announce specific configurations, pricing and availability in the coming months.