Things are a foot at the GPU Technical Conference. And one of those things aims to make developing on GPU via CUDA a better experience.
Nvidia today launched an updated version of their Nsight development suite. This wraps the CUDA development SDK with an Eclipse-based IDE, providing a graphical debugger, inspector and code completion.
The UK Company Enterpoint has started offering the Merrick 6 with 6 Spartan FPGA processors on a single PCI board that runs on a single 12V power supply it
This product should be very popular for applications requiring intense computing from engineering CAD, rendering to bit coin mining.
The Italian Ministry of Education University and Research has announced an upgraded Fermi System that will be installed in Spring and operational in August 2012. This computing power will become available to all European institutions part of PRACE (the Partnership for Advanced Computing in Europe).
there is a growing demand from national and European academia, research institutions and industries for peta-scale computing capabilities.
IBM's Watson that wow'd the public and geeks with it's performance on Jeopardy. Has some what inveitably, been quietly tuned for for use by Wall Street.
Michael Versace, head of risk research at IBM has said, “Analytics is the new core in competitive banking. The ability to efficiently and effectively exploit big data, advanced modeling, text analytics, in memory and real-time decisions across channels and operations will distinguish those that thrive in uncertain and uneven markets, from those that fumble.”
Apache Pig is a high-level procedural language for querying large semi-structured data sets using Hadoop and the MapReduce Platform. Pig simplifies the use of Hadoop by allowing SQL-like queries to a distributed dataset.
Bangalore has been named as the host for the the yet-to-be-named high-performance computing system will be used for genome informatics, geo-science informatics (earth, ocean and atmosphere) and engineering sciences (aerodynamics of planes, development of smart materials and computer-aided drug design).
Rumour has it the supercomputer will deliver a sustained performance in excess of 250 teraflops. This is part of the 12th five-year plan (2012-’17), the Central Government has decided to allot approx 76 million (GBP) (120 million dolars) to propel India into the elite supercomputing club.
The current Indian CSIR-C-MMACS run around 24 T-Flops system. Over the next few years, “CSIR plans to upgrade their supercomputing capacity to 10 petaflops (10 x 1015 ),” said R P Thangavelu, coordinator,
Prof Samir K Brahmachari, director general of CSIR, suggested the Indian facility would play a crucial role in empowering data intensive scientific discovery and promoted data connectivity
“Today, all 40 CSIR labs in India are interconnected using the National Knowledge Network, which enables all scientists to access the supercomputing facility remotely. The new system would enhance the capabilities in areas such as genome analysis, weather modelling, computational fluid dynamics and the like,”
After nearly 6 years the Apache Foundation has released a new version of their flagship project, the Apache Web Server. Although the existing version is stable and is the most popular web server on the planet, new players have been emerging to try and take the crown away from Apache.
The global race to Exascale computing but the worlds governments recieved a boost last week, when the European Union released a statement on it's current approach.
The European Union said last week that it would double it's investment in it's push for Exascale computing, according to an article in ComputerWorld. This would bring the funding to EURO1.2B from the original EURO630 million.
This level of funding now puts the US effort into the shade. Lsat year, the U.S. Department of Energy (DoE) asked for $91 million, but only received $73 million.
University of Edinburgh’s Advanced Computing Facility (ACF) announced plans for the next generation of UK Super Computers for Science. This includes the combined power of the UK's HECToR and BlueGene/Q
The announcement included details on HECToR Phase 3 with a £13.9 million grant from the Engineering and Physical Sciences Research Council (EPSRC).
Professor John Womersley, Chief Executive Officer of STFC spoke of the prominance of High Performance Computing in UK Science.
“Supercomputers are the essential, behind-the-scenes tools that enable modern science. Whether you are analysing climate data from a satellite, designing a new medicine or looking for the Higgs boson, access to high performance computers is vital. These new computers will undoubtedly facilitate breakthroughs across the scientific disciplines, and lead to additional economic and societal benefits for the UK.”
The UK's BlueGene/Q and HECToR are both ranked at 800 Teraflops with approximately one Petabyte of storage.
They perform simulations across a range of scientific disciplines and are funded by four of the UK Research Councils, EPSRC, STFC, NERC and BBSRC.
Have you UK High Performance Computing News? let us know info@scalabiliti.com
Some times good things come in small packages. We've loved the idea of the low cost Pi Linux box since it was announced. The promise of allowing people to create the next generation of technology that is muti-core clusters and smart networked was really compelling.
Raspberry is Pi Foundation's 700MHz ARM11 CPU $25 Linux computer. While much is speculated it is claimed that the Broadcom graphics hardware in the Raspberry Pi offers twice the performance of the iPhone 4S GPU and soundly beats NVIDIA's Tegra 2. What this means for experimental pipe-lined jobs is far from certain. So far the Pi computer has ben demonstrated running smoothly a H.264-encoded 1080p video.
We welcome any further informaiton regarding Pi's usage in experimental/hobiest custering, CUDA and other parallel activities. Email us at info@scalabiliti.com
Source : http://www.raspberrypi.org/
Super dense (Arm) servers is an area we have been tracking for some time. Now SeaMicro steps into the game, kick starting a super dense intel server.
The SM10000-XE uses 1/2 the power, takes 1/3 the space, and delivers 12x the bandwidth of today’s equivalent servers. Packing 64 Xenon (256 2.4 Ghz cores) on a sandy bridge platform.
With funding from Engineering and Physical Sciences Research Council (EPSRC), the University of Manchester is heading up a large interdisciplinary network focusing on numerical algorithms and high performance computing.
Overseen by Professor Nick Higham and Professor David Silvester of the School of Mathematics, the network aims to support interaction and collaboration between UK numerical analysts, computer scientists and developers and users of software and HPC.