Unlimited Scale to Boost Productivity Of High-Performance Linux Clusters

ST. PAUL, MN -- Unlimited Scale, Inc. today announced that it is developing the definitive scalable version of the Linux operating system to support performance-critical application workloads. The product, "Unlimited Linux," will enable customers to manage clusters of tens to thousands of server nodes as a single unified system on which large, complex and dynamic production workloads can be executed efficiently. In addition, Compaq Computer Corporation (NYSE: CPQ) has made an equity investment in Unlimited Scale(TM) and entered into a cooperative agreement to deliver highly scalable Linux systems with the full functionality needed to meet customer demands for high performance technical computing solutions. In connection with this equity investment, Bill Blake, Compaq's Vice President of High Performance Technical Computing, has joined Unlimited Scale's board of directors. The goal of the Unlimited Linux software environment is to provide, on a given hardware configuration, up to twice the system utilization than can be achieved with other Linux releases while significantly enhancing the ease with which large clusters are managed, and delivering increased parallel application performance. Unlimited Scale, Inc. and Compaq Computer Corporation Collaboration Unlimited Scale, Inc. has entered into an agreement with Compaq to collaborate on an optimized version of Unlimited Linux for Compaq's highly scalable high performance technical computing systems. This collaboration will leverage Compaq's experience in developing and delivering Terascale high-performance supercomputers, along with Compaq's leadership in the Linux-server marketplace. "Unlimited Scale brings a wealth of experience in large-scale system software design," said Compaq's Blake. "This level of expertise is needed to extend Linux to efficiently utilize highly scalable computing platforms such as Compaq's AlphaServer and ProLiant Server families. Compaq looks forward to working with Unlimited Scale to meet this important customer need." Unlimited Scale has agreed to collaborate with Compaq to deliver preferential support and optimized versions of Unlimited Linux for Alpha-processor-based Compaq AlphaServer(TM) and Intel(TM)-processor-based Compaq ProLiant(TM) platforms. Unlimited Scale has chosen to support the Compaq AlphaServer platform as their exclusive RISC-processor platform for a three-year period. Delivering Productivity and Manageability on Highly Scalable Linux Systems Clustering is revolutionizing high-performance computing by providing the potential for unprecedented performance at commodity prices. Increasingly, high performance computing users are seeking an efficient large-scale clustered version of Linux to provide both platform independence and access to the growing Linux application base. "I have spoken to a broad range of engineering and scientific users who tell me that they are looking for ways to extend the advantages of Linux to large-scale systems," said Steve Oberlin, President and CEO of Unlimited Scale. "We are uniquely positioned to provide the resource management, scheduling and system administration capabilities that are needed by high-end users. This will be done without impacting our Linux compatibility." Unlimited Linux is expected to be of value to a wide range of users within the national defense and intelligence establishments, drug manufacturers and bioinformatics companies, automotive and aerospace manufacturers, oil companies and exploration contractors, universities and government data centers involved in energy and environmental research, and financial institutions. Unlimited Linux -- The Product The initial product release, based on Sandia National Laboratories' Cplant(TM) clustering software environment, will be available early next year. Unlimited Linux will consist of a standard Linux distribution and a proprietary software layer, which will be commercially licensed and supported. Unlimited Scale plans to provide specific product details during the forthcoming IEEE SC2001 meeting in Denver, Colorado in November 2001. "Cplant was developed at Sandia to run large-scale Linux clusters of up to 2500 nodes," said Bill Camp, Director of Computation, Computers and Mathematics at Sandia National Laboratories. "There is a real need in high performance technical computing for a fully supported product with the capabilities of Unlimited Linux and I am very pleased that Cplant has been chosen as a key component of its core technology." The Linux clusters at Sandia using Cplant software are among the top supercomputing clusters in the world. Cplant uses Compaq AlphaServer DS10 systems to deliver more than two trillion floating point operations per second (2 TeraFLOPS) for use in Sandia's research.