The Week in Review
Here’s a collection of highlights, selected totally subjectively, from this week’s HPC news stream as reported at insideHPC.com and HPCwire.
>>10 words and a link
TotalView now part of Platform OCS distro;
NVIDIA demos CUDA plug-in for MATLAB;
UA deploys FPGA-capable Altix;
Laptop maker Lenovo builds Formula One supercomputer;
>>Saving money on Network.com
There was an interesting article this week about how data processing company Infosolve Technologies has moved its IT service offering from its own datacenters to Sun’s Network.com computers:
When it discovered the Sun Grid Compute utility, Infosolve knew it had found the highly scalable, high-performance solution for which it was looking — and at a cost, by company estimates, $150,000 less per year per server node than traditional hosting services, and $300,000 per year less than building and maintaining an in-house cluster solution.
…What’s so interesting is that utilizing the utility model is the norm at Infosolve, which currently runs data matching jobs across the Sun Grid for every single customer, whereas software-as-a-service programs with many ISVs tend to be among the less-popular options.
Full article at http://www.gridtoday.com/grid/1663866.html.
>>Intel’s Extreme processor
Intel announced new Core 2 Extreme processors this week in desktop and laptop versions. The laptop probably isn’t HPC; the desktop chip might be used in clusters, so here are some stats from the release:
For desktop PCs, Intel is announcing a robust set of new processors, including the flagship Intel Core 2 Extreme QX6850 quad-core processor. Clocked at 3.0 GHz with a new, faster 1333 MHz system bus speed, the Intel Core 2 Extreme QX6850 will be welcomed by game developers looking to deliver new features and performance levels. Intel also announced new Intel Core 2 Duo and Core 2 Quad processors.
Interestingly, Intel is piggy-backing its HPC success onto this announcement:
A year ago this month the company began introducing these innovative and energy-efficient products, and in that year Intel Core microarchitecture and its processors have achieved a number of milestones, including:
…Securing nearly half (224) of the 500 top-ranked high-performance computers as measured by the recently announced Top 500 Supercomputers List (www.top500.org).
>>New SPEC MPI benchmark
The Standard Performance Evaluation Corp. (SPEC) has just released a parallel systems benchmark based on MPI applications:
SPEC MPI2007 provides performance metrics that can be used to compare different hardware architectures (SMP, NUMA, clusters) and interconnects, processors, memory hierarchy, compilers, and MPI implementations. It tests performance based on actual end-user applications, instead of synthetic workloads or parallelized versions of sequential benchmarks.
…Two metrics are provided by SPEC MPI2007: A base metric for users who prefer the relative simplicity of a single-step build process, and a peak metric for those who want to experiment with different compilers and compiler flags to achieve the best performance. Performance measurement is normalized against a reference system to produce a “bigger-is-better” metric. The current reference system is an eight-node cluster of Celestica A2210 systems, each with dual AMD Opteron processors, connected by a TCP (GigE) interconnect.
MPI2007 complements other HPC benchmarks already available from SPEC: HPC2002 and OMP2001 (details at spec.org). The benchmark is available now, but you’re going to have to pony up $800 for it. Details at http://www.forbes.com/businesswire/feeds/businesswire/2007/07/17/businesswire20070717006085r1.html.
John West summarizes the headlines in HPC every day at insideHPC.com, and writes on leadership and career issues for technology professionals at InfoWorld and on his own blog at onlytraitofaleader.com. You can contact him at firstname.lastname@example.org.