Thursday, 2 June 2016
Thursday, 26 March 2009
Earth Hour Saturday
Join Quadrics, and vote with your light switch at 8.30pm on 28 March.
Image courtesy of Earth Hour UK http://earthhour.wwf.org.uk/how_you_can_help/earthhourcorporate/
Saturday, 7 March 2009
Quadrics 10 GigE Switch Case Study - GreenLight Project by UC San Diego
Getting more computational work done for fewer watts of energy
Evolution of Information Technology and the increasing energy demands that go with higher computational power have historically been an obstacle to sustainability. Recent IDC research shows that over the past ten years the average cost to power and cool an installed base of servers has doubled . A project run by the University of California, San Diego aims to measure the energy efficiency of new technologies such as 10 Gigabit Ethernet (10 GigE) and their impact on cluster capacity and efficiency - and reduction of energy costs.
How to enable green IT computer scienceA green strategy requires various steps: monitoring and controlling the use of electricity of network-connected devices, design innovative ways to reduce energy consumption and then more monitoring and measuring to refine the process. Academia and enterprise often collaborate in this research, looking for ways to get more computational work done for fewer watts of energy. This is the scope of GreenLight, a project run by the scientists and engineers of UC San Diego, one of the greenest universities in the US. When it is built, the GreenLight Instrument will measure, monitor and help optimize the energy consumption of large-scale scientific applications from many different disciplines. What makes the project exemplary is that the cluster used to run the measurements and the monitoring has been designed to use less energy than most data centers: GreenLight consists of an eco-friendly Sun Modular Datacenter container with 8 racks of servers. When fully populated, the servers will connect to 7 edge switches via 1 Gbps links and the edge switches will connect with 10 GigE uplinks to the Quadrics TG201-XA switch, which maintains low latency and high bandwidth and ensures the end users quality of service. The immediate advantage, as Tom DeFanti, GreenLight principal investigator, explains, is that "GreenLight uses 10 Gbps over dedicated optical fiber links so end users move their clusters out of their faculty closets and into much greener configurations".
Evolution of Information Technology and the increasing energy demands that go with higher computational power have historically been an obstacle to sustainability. Recent IDC research shows that over the past ten years the average cost to power and cool an installed base of servers has doubled . A project run by the University of California, San Diego aims to measure the energy efficiency of new technologies such as 10 Gigabit Ethernet (10 GigE) and their impact on cluster capacity and efficiency - and reduction of energy costs.
How to enable green IT computer scienceA green strategy requires various steps: monitoring and controlling the use of electricity of network-connected devices, design innovative ways to reduce energy consumption and then more monitoring and measuring to refine the process. Academia and enterprise often collaborate in this research, looking for ways to get more computational work done for fewer watts of energy. This is the scope of GreenLight, a project run by the scientists and engineers of UC San Diego, one of the greenest universities in the US. When it is built, the GreenLight Instrument will measure, monitor and help optimize the energy consumption of large-scale scientific applications from many different disciplines. What makes the project exemplary is that the cluster used to run the measurements and the monitoring has been designed to use less energy than most data centers: GreenLight consists of an eco-friendly Sun Modular Datacenter container with 8 racks of servers. When fully populated, the servers will connect to 7 edge switches via 1 Gbps links and the edge switches will connect with 10 GigE uplinks to the Quadrics TG201-XA switch, which maintains low latency and high bandwidth and ensures the end users quality of service. The immediate advantage, as Tom DeFanti, GreenLight principal investigator, explains, is that "GreenLight uses 10 Gbps over dedicated optical fiber links so end users move their clusters out of their faculty closets and into much greener configurations".
Labels:
10 Gigabit Ethernet,
Calit2,
case studies,
cost effective,
DCIE,
Green IT,
PUE,
Quadrics,
sustainability,
UC San Diego
Monday, 2 March 2009
HPC Asia - Taiwan
HPC Asia is the HPC conference for APAC, it occurs every 18 months in a different country of the region and it is currently being held at the Grand Hi-Lai Hotel, Kaohsiung, Taiwan.
It's always good to see familiar faces and Quadrics customers among the speakers invited, to name just a few: Jack Dongarra from the University of Tennessee, Peter Arzberger from the University of California, San Diego, Dr. Zhiwei Xu from the Chinese Academy of Science, Mark Seager from LLNL.
For those interested to see the whole programme please click on this link.
Quadrics local sales representatives are attending, if you want to get in touch with them feel free send an email to sales@quadrics.com with your mobile number and the best time you'd like to meet and we'll follow it up.
Labels:
APAC,
Asia,
Chinese Academy of Science,
HPC,
HPC Asia,
Jack Dongarra,
LLNL,
Mark Seager,
Taiwan,
UCSD
Thursday, 26 February 2009
QsNetIII, An HPC Interconnect For Peta Scale Systems
Check out this SlideShare Presentation:
Friday, 20 February 2009
Parallel Performance Wizard (PPW), performance analysis tool for PGAS
Given the complexity of parallel programs, developers often must rely on performance analysis tools to help them improve the performance of their code.
HCS Research Laboratory have written a very interesting whitepaper, Parallel PerformanceWizard: A Performance Analysis Tool for Partitioned Global-Address-Space Programming, which includes how one would use PPW in the analysis and optimization of PGAS applications by presenting a small case study with a Quadrics QsNetII cluster using the PPW version 1.0 implementation.
PPW is known to work on many platforms and it has been heavily tested on Linux with the Berkeley UPC compiler and the GCC UPC compiler.
Parallel Performance Wizard (PPW) is a performance analysis tool designed by the University of Florida High-performance Computing and Simulation (HCS) Research Laboratory for UPC and SHMEM programs.
Sharing best practice:
HCS Research Laboratory have written a very interesting whitepaper, Parallel PerformanceWizard: A Performance Analysis Tool for Partitioned Global-Address-Space Programming, which includes how one would use PPW in the analysis and optimization of PGAS applications by presenting a small case study with a Quadrics QsNetII cluster using the PPW version 1.0 implementation.
PPW is known to work on many platforms and it has been heavily tested on Linux with the Berkeley UPC compiler and the GCC UPC compiler.
PPW version 2.0 was released on 01/21/2009.
To download please go to: http://ppw.hcs.ufl.edu/download.html
Image from: "Parallel PerformanceWizard: A Performance Analysis Tool for Partitioned Global-Address-Space Programming", Hung-Hsun Su Billingsley, M. George, A.D. Dept. of Electr. & Comput. Eng., Univ. of Florida, Gainesville, FL, 2008
Thursday, 19 February 2009
Quadrics Supported Linux Distributions
There are a wide variety of Operating Systems, Kernels, and Hardware Platforms in use by Quadrics' customers in the field.
The page from this link describes the configurations that are officially supported by Quadrics.
Quadrics supports the currently shipping release of both Red Hat Enterprise Linux (RHEL) and Novell SUSE Linux Enterprise Server (SLES). As of January 2009, these are RHEL 5 and SLES 10.Quadrics also supports the previous major release from Red Hat and Novell (currently RHEL 4 and SLES 9).
Quadrics support for SUSE Linux Enterprise Server 11 (which will be released in 2009 or 2010) will be announced on this page.
Support for other Operating Systems will be considered where there is a business case.
http://www.quadrics.com/Quadrics/QuadricsHome.nsf/DisplayPages/85F1275AD2C30C6380256FBA003C61FB
Subscribe to:
Posts (Atom)