NextIO is a company with a cool offering that allows you to pool your network ports, disk drives, GPUs, and other IO devices into a box that can be connected via PCI Express to multiple servers. As the company says, this lets you separate decisions about IO from decisions about compute. A use case? Rather than adding GPUs to just a few servers in your small cluster you could add them in a NextIO device that would make them available to all the servers. The point is not that all the servers would use them simultaneously (that would be bad), but rather that you wouldn’t have to decide a priori which servers would get the GPUS, allowing you to avoid idle resources held “just in case” a GPU user showed up.
This week they announced a partnership with IBM that puts their gear into IBM clusters for users that can take advantage of it
NextIO, a premier provider of next-generation I/O solutions, today announced it is working with IBM to offer customers integrated cluster solutions that incorporate NextIO technology, with availability in 2010. The solution will enable reconfigurable and on demand GPU compute capabilities for IBM iDataplex customers. The announcement was made at the Supercomputing 2009 show in Portland, Ore.
…The GPU virtualization solution will offer the ability for a single IBM iDataPlex™ server to access from one to eight double-wide GPUs or up to 16 single-wide GPUs in the appliance. Users can quickly and easily enable more or less GPU resources on demand, depending on application requirements. Each iDataPlex rack supports 10 GPU appliances providing up to 160 GPUs and over 80TFlops of compute processing per rack.
If you were at SC09 this year, you might be interested in knowing that a number of those exhibiting were a lot closer to home than most of us were. Local companies included big companies with Oregon offices, like Intel and Sun, and companies headquartered there like compiler maker The Portland Group, RNA networks, and VirtenSys.
Intel’s Software Tools group has announced a cloud-based “scalability service.” The Intel Parallel Universe Portal is an on-demand cloud-computing analysis tool that tests 32-bit Windows-based parallel applications. The service lets software developers assess how their applications will perform on a number of multicore processor configurations — 1, 2, 4, 8, or 16 hardware threads — without having large multicore systems in-house.
The Appro Xtreme-X Supercomputer is a highly scalable architecture that groups high performance servers together into a unified, fully integrated Scalable Unit (SU) that can be provisioned and managed as a stand-alone supercomputer. The Appro Xtreme-X Supercomputer is based on the Quad-Core Intel Xeon Processor offering interconnect fabric options for single rail InfiniBand Network for price/performance or dual rail configuration for performance/availability computing.
Platform Application Center is available as an optional capability for both
I was at former Vice President Al Gore’s keynote talk on Thursday at SC09. I had great seats; four or five rows back and off to the left. Fun fact: this was the second time he’s been at the show — evidently he attended as a senator in the early years of the conference. I wouldn’t have thought I was the kind of person that was interested in being in the room with someone that famous at least partially because he is famous, but I am. There you go.
Penguin Computing, experts in high performance computing solutions, today announced that Tesla GPU compute nodes are available in its Penguin on Demand (POD) system. Tesla equipped PODs will now provide a pay-as-you-go environment for researchers, scientists and engineers to explore the benefits of GPU computing in a hosted environment.
Sabalcore Computing Inc. has selected Scalable Informatics as their primary storage vendor to provide high performance storage capabilities to Sabalcore customers. “We chose Scalable Informatics because of their depth of experience, excellent service, and extremely reliable products,” said John Van Workum, President of Sabalcore Computing.


