Because I have the ability to peer into the future, I know that on May 29th Linux Networx will announce
the immediate availability of it’s second generation LS-V High Performance Visualization System, providing the ability to render very large simulations in real time and delivering insightful results quicker and more efficiently than ever before. The enhanced LS-V delivers a system experience that uniquely balances ease of use with compute, interconnect, high performance I/O, and visualization based upon real-world, proven high-performance computing (HPC) expertise.
What’s new with 2.0?
Second-generation LS-Vs feature a range of updated AMD Opteron Dual-Core CPUs, InfiniBand interconnects, fully integrated data I/O subsystems, and the optional Linux Networx GPFS high performance file system. The LS-V is configured in turnkey, production-ready systems ranging in size from 3 to 31 visualization nodes, fully scalable from individual users through large corporate visualization centers.
I am a bit surprised that in the press release they include the following statement
The innovative technology that forms the core of the LS-V has been previously deployed in very large Linux clusters, including systems for the Army Research Laboratories…
I’m surprised because the hardware they recently installed at ARL took almost a year to get going anywhere close to operational, and when last I checked still wasn’t working properly. That purchase has been plagued by severe technological and execution problems — definitely not a success story. I’m well aware that large systems have problems, believe me. But the fact that they are referencing ARL as an exemplary success reflects that the company may not have learned the lesson it needed to.