Platform Computing (Platform), a provider of workload and resource management software for clusters, grids and clouds, has entered into an agreement whereby Platform has acquired HP-MPI (Message Passing Interface) from HP. This acquisition supports Platform’s strategy to extend its offering in High Performance Computing (HPC) management software with development of a MPI product.
Under the terms of the deal, HP will continue to sell and market HP-MPI under the Platform MPI brand. Platform has purchased the technology assets and intellectual property associated with HP-MPI software. In addition, Platform has hired key members of the HP-MPI development team.
HP-MPI is an integrated implementation of MPI, a message passing protocol used by resource intensive applications to run parallel processes over multiple nodes in a computing cluster. Platform currently has an MPI product (formerly Scali MPI) and a key objective for the company has been securing broader adoption of Platform MPI by more independent software vendors (ISVs).
HP-MPI is used and distributed by more than 30 ISVs, making it an adopted MPI software across industries. Platform’s plan is to combine the performance of its existing MPI technology with HP-MPI’s ISV acceptance to create a single product named Platform MPI.
HP-MPI and Platform’s current MPI products will continue on current release cycles and be maintained to support existing customers for an interim period. In addition, Platform will continue support for legacy versions of each product even after the release of a combined offering.
Platform said that it will continue to develop and enhance the MPI product by working with HP. HP will resell Platform’s MPI technology and Platform will build upon HP’s marketing and product support for ISV’s who have standardised their parallel applications and cluster support around HP-MPI.
The company has been enhancing its portfolio in other HPC areas also. Last week, it released new GPU kits to support NVIDIA’s CUDA-enabled GPUs, which makes it easier for administrators to build, provision, manage and schedule workloads on NVIDIA GPU-based HPC clusters.
