HPCwire is carrying a release from HPC hardware startup Liquid Computing about its new partnership with Scali Scali and Liquid Computing announced an OEM agreement supporting Scali MPI Connect on Liquid Computing’s LiquidIQ server. Scali MPI Connect is a high performance, scalable and fully supported implementation of the industry standard Message Passing Interface (MPI) which […]
Search Results for: iq
Canadian-based Liquid Computing is announcing today that its LiquidIQ server has demonstrated world leading results for the STREAM Benchmark for 4 socket Second-Generation AMD Opteron processor-based servers. Using the latest AMD Opteron processors Model 8220 from AMD, LiquidIQ’s Compute Modules can provide more than 20,000 Megabytes per second on all of the STREAM Benchmark Bandwidth […]
Liquid Computing announced last week that it has been working since January with Dalhousie University’s Faculty of Computer Science in Nova Scotia to advance the state of the art in parallel processing for very large databases: The partnership project…has already demonstrated the dramatic performance advantages of applying new computing architectures to data warehousing problems. …“We […]
I’m happy to see Liquid talking publicly about performance results on the HPCC benchmarks. But their marketing seems unnecessarily aggressive and a little too enthusiastic. The title of their press release is “Liquid Computing throws down the gauntlet with HPC Challenge results,” subtitled “LiquidIQ beats Cray XT3’s posted latency performance.” They go on to cite […]
Liquid Computing is announcing today its first customer acceptance of the LiquidIQ platform. The system was placed at the Army High Performance Computing Research Center (AHPCRC). No details in the release on the configuration (if you’re a reader with details, let me know). From Liquid’s web site: Liquid Computing Inc. is first to deliver a […]
I’m very happy to see a new crop of hardware vendors popping up that are doing new things. Companies like SiCortex and Liquid Computing. Mind you I’m not assessing the technology (yet: but I do have in depth assessments of these offerings in the works), I’m just glad that somebody out there is doing more […]
This week we look at various attributes including how easy it is to scale Lustre file systems. The inherent scalability of Lustre aggregates storage capacity across many servers. I/O bandwidth also scales as more storage servers are added, and can be dynamically adjusted as needs change and demands for more storage capacity and bandwidth grow.
As the countdown to Exascale continues, Exascale-like storage problems are already showing up in today’s massively parallel, heterogeneous HPC systems. Historically, storage and I/O have kept pace with growing system demands, but, because of the limitations of spinning media and the cost of solid state storage technologies, storage performance improvements have come at a disproportionately higher cost and lower efficiency than their compute counterparts.