X
Business

IBM announces big data focused Smarter Computing" initiative

IBM is focused on bringing its technical computing products to a larger audience. This announcement also offers hints as to IBM's future plans to bring intelligence to each component of a distributed computing environment.
Written by Dan Kusnetzky, Contributor

IBM just launched a Smarter Computing Initiative designed to help its customers use Big Data processing and storage techniques. It was a large announcement containing many disk and tape storage product enhancements, additions to IBM System x Intelligent Cluster and new versions of several Platform Computing products. Platform Computing, by the way, was recently acquired by IBM. I'm going to focus on the selected products mentioned in this launch.

What IBM has to say about selected products in the Smarter Computing Initiative

  • The IBM Platform Symphony family, a grid manager that is now integrated with the MapReduce software framework to provide faster throughput and performance for demanding analytics and big data workloads in a single grid environment. Platform Symphony’s resource sharing model makes it cost-efficient for clients to expand their analytics environment as needed.
  • The IBM System x Intelligent Cluster integrated with IBM Platform HPC software to simplify cluster deployment, deliver results more quickly and improve productivity so clients can focus on research and analysis instead of managing their IT infrastructure.
  • The High Performance Computing (HPC) Cloud portfolio from IBM expanded with the new IBM Platform Cluster Manager offering as well as IBM Platform LSF to provide clients with a shared pool of cloud resources available from anywhere, making it easy to create and manage HPC clouds. The new IBM Platform Cluster Manager enables clients to self-provision clusters in minutes and automatically, dynamically manage cluster environments that include both IBM Platform Computing and non-IBM workload managers.
  • IBM Platform LSF family – IBM Platform LSF is a comprehensive set of intelligent, policy-driven workload management and scheduling tools that can be used to manage high performance workloads across a distributed, virtualized IT environment with up to 100 percent utilization to help keep costs low.
  • IBM General Parallel File System (GPFS) – GPFS now includes Active File Management (AFM) software to provide fast, trusted access to unstructured data regardless of where the data resides so it can quickly be turned in to insight.
  • IBM System x iDataPlex dx360 M4 – The latest iDataPlex system can double performance on selected workloads with the latest Graphics Processing Units (GPUs) from NVIDIA, increases maximum memory to 512GB and – with the new slotless FDR (Fourteen Data Rate) Infiniband adapter – provides superior performance and flexibility for technical computing environments.

Snapshot Analysis

Although a number of new, improved software, storage and system products were announced, much of the story revolves around the center of better and more intelligent integration in the vast portfolio of IBM products that can be used to support technical computing. This integration effort extends to bringing the excellent Platform Computing products into the IBM family.

The announcement also hints at things to come. IBM clearly is doing its best to bring distributed, local processing to its technical systems, storage systems and even into its distributed processing software products, such as the IBM Platform LSF Family and IBM General Parallel File System (GPFS). Why? To reach towards its goals of offering Exascale computing (see Is workflow analysis better than relying on benchmarks? ) for other clues IBM has offered about its thinking.) I expect to hear more about adding some level of intelligence to storage, networking, virtual machine management and several other important topics over time.

The Platform Customers have been using Symphony and LSF to support technical computing, modeling for commercial workloads and Big Data applicaitons for years. Unfortunately, very little of that success has shown up on the industry radar screen. That's really unfortunate because the customers I've spoken with find the tools very powerful.  Now that IBM's Platform Computing group has made it possible for users of MapReduce to manage their efforts using Platform's tools, it is very likely that we're going to hear more about Platform's successes.

This was an interesting launch that offered many useful enhancements and additions to the IBM portfolio. What was more interesting was the hints about what IBM is planning to bring to the market in the future.

Editorial standards