IBM announced that it has entered into an agreement to acquire one of the pioneers in processing virtualization, Platform Computing. This move will extend IBM's reach into the areas of technical computing and, in all likelihood, certain areas of cloud computing as well.
I've been following Platform Computing for ages. While I was at DEC, Platform Computing was a valuable partner. While I was at IDC, Platform was a client. I've often thought that if the company's technology could be marketed properly, it would be in use in more places.
What does Platform Computing do?Platform Computing offers a number of products that help organizations manage, provision, automate and orchestrate a large number of systems and focus them on a single task or a small number of tasks. Platform has been one of the forerunners in the areas of clusters, compute grids or, most recently, technical computing clouds.
Here are a few of the products Platform offers today:
- Platform ISF — A tool designed to help organizations create, provision and manage self-service applications based upon virtual servers hosted by a number of different hypervisors/virtual machine monitors. It is flexible enough to allow management of some or all of these resources to be delegated to different groups within the organization. Applications are managed using policies that can offer the organization a great deal of control, enhance the ability of the organization to maintain regulatory compliance, and increase levels of security.
- Platform LSF — A tool designed to manage workloads across multiple platforms. As with Platform ISF, workloads can be managed based upon policies. At this point, the tool has the ability to manage up to 1.5 million pending jobs (which could be made up of multiple processes) and control execution using more than 100,000 cores.
- Platform Symphony — A tool designed to manage and orchestrate distributed application services. These services could be executing on multiple different operating systems, supporting different hardware architectures and still be seen as a single computing resource.
- Platform Cluster Manager — A tool designed to make management of clusters easier.
- Platform MPI — An implementation of the message passing interface (MPI) standards allowing Linux and Windows systems supporting a number of workload managers to be integrated to support complex compute-intensive tasks.
- Platform HPC — A high performance computing manager offering integrated management capabilities for high performance computing clusters.
Snapshot analysisPlatform Computing is all over managing clusters, grids and their cloud computing counterparts. It's tools are powerful, comprehensive and offer organizations the ability to utilize complex system configurations. This is both a strength and a weakness. The products are very powerful, but often seem a bit overwhelming to the uninitiated. The appearance of complexity has caused some to consider technology that was much less capable, but appeared to be more tightly focused on the needs of a single market.
I believe IBM will be able to take this technology, integrate it into its "Smarter Computing" marketing programs and introduce many organizations to the benefits of harnessing together the power of a large number of systems to tackle very large and complex workloads.
This is a good match.