by Anton Shilov
09/16/2013 | 11:05 PM
IBM has introduced the NeXtScale System, a flexible computing platform providing three times as many cores as current one-unit rack servers, making it ideal for the fastest growing workloads such as social media, analytics, technical computing and cloud delivery.
The rapid adoption of these workloads and delivery models is putting increased demands on data centers, and clients are looking for new technologies that meet those demands with the highest performance and the lowest power consumption to drive high efficiency and lower costs. NeXtScale is the latest addition to IBM's x86 portfolio, designed specifically to run those applications with the power of a supercomputer in any data center, via a simple, flexible and open architecture that will support options for compute, storage, and graphics processing acceleration.
IBM NeXtScale server system incorporates up to 84 x86-based systems and 2016 processing cores in a standard EIA 19” rack, allowing easy integration into any data center. It uses industry-standard components including I/O cards and top-of-rack networking switches for flexibility of choice and ease of adoption. IBM also provides a powerful software stack to run on top of NeXtScale, including IBM general parallel file system, GPFS storage server, xCAT, and platform computing, providing powerful scheduling, management and optimization tools.
NeXtScale supports the industry's fastest x86 processors (Intel Xeon processor E5-2600 v2) and DDR3 1866MHz memory, and can be integrated into data centers or departmental IT closets, including those running at 100V-127V power inputs. NeXtScale is approved for operation in higher-temperature data centers (up to 40C/104F degrees), reducing cooling requirements and further lowering operational costs for users. NeXtScale’s native expansion concept allows users to add common functionality such as storage, graphics acceleration or co-processing, either at the time of shipment or in the future. Available with NeXtScale are new solution starter kits that make it easier for users to configure many common departmental HPC and small cloud solutions, such as Ansys, MPI-BLAST, and OpenStack.
The result is a single architecture based on open standards that delivers high performance and high efficiency, and that is designed to blend seamlessly with clients' data centers, current practices and x86 tools. NeXtScale's design can help clients better manage their operations and capital expenditure budgets by allowing them to maximize the compute power in a minimum amount of space in their data centers. Because of its standard-rack form factor and broad use of industry-standard components, IBM Business Partners can now deliver IBM's high performance computing technology to a broad range of users.
IBM NeXtScale is designed for:
"By delivering high flexibility, ultra performance and great efficiency in an industry-standard format with a powerful software stack, NeXtScale provides our clients with a versatile new data center solution that is easy to qualify, acquire and deploy. The introduction of NeXtScale and the other offerings and enhancements we're announcing today illustrate our continuing commitment to providing high-value, high-performance x86-based solutions to our clients," said Adalio Sanchez, general manager at IBM System x.
Built for flexibility, NeXtScale can be purchased as a single node, an empty or configured chassis, or in full racks as a complete pre-tested IBM Intelligent Cluster solution, delivered fully configured and ready to power on. With Intelligent Cluster, NeXtScale arrives at the client location racked, cabled, and labeled with user-supplied naming, with pre-programmed IMMs and addresses, and with burn-in testing completed at virtually no additional cost. This can reduce time from arrival to production readiness by 75%, while also substantially reducing packaging waste.