Red Hat follows hot on Microsoft’s HPC heels

gallery

Red Hat has introduced its first integrated Linux-based, high-performance computing (HPC) platform.

But the market leading commercial Linux vendor is playing catch up to the likes of Novell and its SUSE Linux Enterprise Server, and Microsoft, which only last week launched an HPC offering based on its Windows Server 2008 product.

The Red Hat HPC stack has been made available globally and offers its Red Hat Enterprise Linux (RHEL) 5.2 operating system (OS) and an original equipment manufacturer version of Platform Computing’s Open Cluster Stack 5 to provide an open source cluster management toolset. This includes device drivers, a cluster installer, a resource and application monitor, interconnect support and Platform’s job scheduler, Lava.

The Linux vendors had made an earlier version of this HPC stack available late last year in the US only. But, where Red Hat had been unwilling to talk pricing last November, it now says this generally available offering will cost $249 (£140) per server node in a cluster.

Compared to the US pricing of Microsoft’s HPC product of $475 (£257) per node, or even the cost of a basic subscription to Red Hat’s standalone RHEL 5 Server running on two sockets of $349 (£197), the new Red Hat HPC product is aggressively priced.

Like Microsoft, Red Hat claimed its launch would take the pain and cost out of deploying HPC and grid infrastructures for smaller to mid-sized businesses (SMBs) particularly.

In fact, Red Hat claimed “customers of all levels and workloads are offered the benefits of high-performance computing through an easy-to-install, end-to-end solution that can be deployed in under one hour”.

Whether that’s true remains to be seen, but that didn’t stop Scott Crenshaw, vice president of the Platform business unit at Red Hat throwing down the gauntlet to its competitors when he said: “We’re enabling our customers to focus on their business goals and competitive advantage without needing to worry about the challenges of deploying and managing their HPC cluster – we’re taking care of this for them.”

Email to a friend

Print this page