Here's a collection of highlights, selected totally subjectively, from this week's HPC news stream as reported at insideHPC.com and HPCwire.
>>Energy in the news this week
Much of this week's news was dominated by word of HPC vendors and service providers starting to do their part to keep the Northern hemisphere from melting and sliding around the bottom of the planet.
At the beginning of the week The Green Grid announced that industry heavyweights Microsoft and Intel joined their ranks, beginning a week of energy-awareness campaigning by the organization. This week marked completion of the organization's formation, and they also announced membership in their board (AMD, APC, Dell, HP, IBM, Intel, Microsoft, Rackable Systems, SprayCool, Sun Microsystems, and VMware).
Later in the week The Green Grid released three whitepapers addressing why it is so important to define and propagate the best energy efficiency practices in datacenter operation, construction and design; a framework for improving the efficiency of both new and existing data centers; and the use of the Power Usage Effectiveness (PUE) metric, along with its reciprocal the Data Center Efficiency (DCE) metric.
In a move that was probably timed to keep Intel on energy message this week, Intel started talking last Friday about the environmental virtues of quickly replacing old servers. Speaking at a conference on data center design last week, Pat Gelsinger, Intel's senior VP in charge of its digital enterprise group, said, “A six year-old server takes up valuable resources that could be better used, so we have accelerated our refresh rate. Refreshing one data center gave us three times the performance for only four per cent more space utilized.”
>>House begins to move on new HPC legislation
The US House of Representatives held hearings on Wednesday to mark up (among others) the HPC R&D Act. H.R. 1068 amends the High Performance Computing and Communications Act of 1991 which established the Networking and Information Technology Research and Development program. The 108th and 109th Congresses both considered similar legislation and failed to act. The bill made it out of committee and now proceeds (hopefully) to a vote.
“This version differs from the most recent attempt (H.R. 28, introduced in the 109th) in that it doesn't attempt to authorize specific agency activities…. First, it directs the Director of the White House's Office of Science and Technology Policy to develop and maintain a research, development, and deployment roadmap for the provision of federal high-performance computing systems. Second, there's an explicit requirement that the President's advisory committee for IT (now a responsibility of the PCAST) review not only the goals of the NITRD program but the funding levels as well and report the results of that review to Congress every two years.”
>>IBM develops new effort to push HPC into the mid-market
IBM was talking this week about a new concerted effort to drive HPC down into the mid-range by making it easier for non-traditional HPC users to deploy small(ish) clusters. This is primarily a marketing push, addings SKUs of cluster configurations prebuilt to meet the needs of specific industries and developing marketing relationships that will help its partners support the industry verticals.
But there was some interesting news aimed at crossing pretty big gaps. The first is that IBM is building four benchmarking and tuning centers aimed at M$oft's CCS 2003. These will be key to convincing new users to buy and then getting them up and going once they do buy. The new centers are located in Poughkeepsie, N.Y., Raleigh, N.C., Beaverton, Ore., and Montpelier, France.
The second important step they are taking is that they are using their deep pockets to make systems available for users to test their applications before they buy. From their release: “Customers interested in moving to a clustered environment can quickly test-drive their HPC applications in IBM's Deep Computing Capacity on Demand centers. With access to over 20,000 processors, customers large and small can tap into IBM Systems to help accelerate time to market and improve quality while helping to keep fixed costs to a minimum.”
I think this is all good news: I'm strongly in favor of anything that gets more people using HPC via the million monkeys theory.
>>U.N. and Silicon Valley team up to boost tech in developing countries
Members of a special United Nations group met this week with Silicon Valley leaders to explore how the technology industry, government and non-governmental organizations (NGOs) can come together to bolster development around the world.
Intel Corporation and the U.N.'s Global Alliance for Information and Communications Technology and Development (U.N. GAID) jointly organized the summit. Intel Chairman Craig Barrett, who also chairs the U.N. GAID, said the meeting provided a critical platform for sharing ideas and forging partnerships among prominent members of government, the international development community and Silicon Valley's businesses, academia and the venture capital industry. Representatives from more than 30 countries attended the summit.
The opening remarks by Barrett, Hamadoun Touré, secretary-general of the International Telecommunications Union, and Farrukh Qayyum, Pakistan's secretary of IT and Telecom, are available at un-webcast.edgesuite.net/un/index.htm. The webcast will be available for replay through March 7. Podcast interviews related to this event are also available on the “About Intel” channel intelpr.feedroom.com. More information about the U.N. GAID is available at www.un-gaid.org.
For a striking image of the worldwide digital HPC divide, check out the Google map at http://www.top500.org/lists/2006/11/top100map showing the physical location of all the sytems in the Top500 . Note that there is exactly one (1) HPC system south of the equator on this list, and zero Top500 systems on the _continents_ of Africa and South America.
IDC released its Q406 Worldwide Quarterly Server Tracker study this week. Their analysis shows HP leading the HPC vendor pack again at the pole position in revenue for the HPC market in 2006 (33 percent market share). This marks the fourth consecutive year of HP revenue leadership in the high-performance computing market (more at http://www.hp.com/hpinfo/newsroom/press/2007/070226b.html).
IBM was talking about a scholarly article in Physical Review Letters (Volume 98, Number 3) detailing the science behind their use of their Blue Gene supercomputer in Zurich to model the material properties of hafnium silicates as they try to shape the behavior of the new material that both IBM and Intel are hanging their next generation chips on.
SGI announced this week that the HPC resources at the Air Force's Aernautical Systems Center (a DOD HPC Modernization Program center) are getting a boost from an Altix 4700 with over 9,000 Itanium 2 cores.
QLogic announced Wednesday that the first iSCSI HBA functionality was accepted into the upstream Linux kernel, marking the first time iSCSI functionality ships with Linux out of the box.
NCSA announced that it is doubling its compute capability to 146 TFLOPS with the addition of a 1,200 quad-core Xeon processor Dell blade system and a 1,040 dual-core Intel system. Both are expected be on line by late spring, according to NCSA.
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 http://onlytraitofaleader.com/. You can contact him at [email protected].
www.onlytraitofaleader.com Leadership and career skills to help scientists, engineers, and technologists find success doing what they love to do. No time to keep up? Subscribe to the RSS feed!