ScaleMP Announces Appro GreenBlade Support

scalempScaleMP, today, announced support for Appro’s line of GreenBlade servers with Intel Xeon processors.  With ScaleMP’s vSMP technology, one can aggregate up to 16 GreenBlade servers to deliver 4TB of memory and 128 cores in a virtual SMP.

The innovative GreenBlade System is flexible, easy to manage and delivers outstanding performance in a modular solution building block that can be tailored for a range of processing, networking, storage or I/O needs,” said John Lee, VP of Advanced Technology Solutions for Appro. “By combining the Appro GreenBlade System with ScaleMP’s innovative vSMP Foundation, we are able to aggregate multiple gB222X blades into a single shared memory virtual system that can support large memory and large core count workloads in addition to parallel workloads, while simplifying the installation and management of the system.”

For more info, read the full release here.

Comments

  1. This vSMP Foundation seems a really neat product, at least on paper. Except a few benchmarks and some buzzword-compliant marketing technobabble, concrete information about ScaleMP’s product (how does it work, does it survive a node failure, what about a feature list?) is pretty much non-existent.

    Do insideHPC readers (or writers) have any first-hand experience with it?

  2. Kilian,

    vSMP Foundation provides several levels of resiliency: redundant (active-active) InfiniBand support, auto restart on node failure, fault-tolerant mode (skipping nodes in case of failed hardware), and more. Note that node failure will lead to OS failure, so auto restart and fault-tolerant mode are extremely important.

    Will be happy to provide more info! Have you got our white paper? Please go to http://www.scalemp.com/info-request, and request white paper. It will clarify many questions.