This week on the Sun HPC Community Portal

Check out these items of note this week on the Sun HPC Community Portal:

Videos:

* Andy Bechtolsheim on Open Storage Townhall
* YouTube: New Sun x64 Systems with Quad-Core Processors
* Future Filesystems: Intelligent, Object-based Storage

Blogs:

* Andy Bechtosheim to Keynote Sun HPC Consortium in Dresden
* Growing Flowers with Datacenter Heat
* Announcing Grid Engine 6.2 Beta Binaries

Events:

* Design Automation Conference, Los Angeles, CA, June 8-13

* Sun HPC Consortium, Dresden, Germany, June 15-17

* International Supercomputing Conference ISC08, Dresden, Germany, June 17-20

From LOST to FOUND: Finding The Island Using Satellite Images and Supercomputers

As the hit ABC television show Lost begins to wrap up its 4th season, a few of us here at Sun started thinking about the ways HPC could be used to crack one of the show's biggest mysteries -- finding “The Island.”

One thought was that perhaps those looking for islands in the middle of Pacific Ocean would be able to find “The Island” by using high resolution satellite images.

High end image processing requires tremendous computing power. To find patterns and information in very large resolution images that humans cannot see requires lots of processing power. Using a cluster of Sun Fire x64 systems very, very large images can be split up across many computing elements to search for patterns. The sheer size of each image, multiplied by a number of images taken of a particular area (different times of day), requires significant storage capacity too. At that scale, a supercomputer could provide quick access and analyze the data.

And its not a pipe dream. The University of Arizona is researching how to use Sun's HPC systems on very high resolution images when searching for signs of life and other objects on distant planets and moons.

The Oceanic Six, Ben, Charles and Penny Whidmore are spending millions of dollars trying to get to the island – maybe all they really need is some Supercomputing horsepower!

AMD Redraws Server Processor Roadmap

Michael Feldman of HPC Wire has this story about recent changes to the AMD processor roadmap:

"Further out, AMD has a chance to catch Intel with the third generation Opteron processors. Two new offerings, the six-core "Sao Paolo" and the twelve-core "Magny Cours," will be developed for a new platform -- what the company's calling Socket G43. Taking a cue from Intel's multipackaging approach, the twelve-core Magny Cours is actually two six-core chips jammed into the same socket." Full Story

If you haven't been to the HPC Wire site lately, you should check out their new design. The watercooler has even made their blogroll!

Cloud Computing on NPR

The term "Cloud Computing" reached a milestone this week as it was featured on NPR's Morning Edition:

"Every Monday, the 'Morning Edition' has a technology theme. The Cloud Computing segment was high level and aimed primarily at a non-tech audience. I always find it hard to answer the question ‘what is Cloud Computing?’ as there are so many different definitions. Regardless, it was a great chance to talk about an exciting technology and highlight the need for a real security conversation between the providers and people interested in IT security - the primary reason why I created cloudsecurity.org." Full Story

NVIDIA shaking up parallel programming world

NVIDIA’s CUDA system, originally developed for their graphics cores, are finding migratory uses into other massively parallel computing applications:

"The argument there is that while it’s not as efficient as it could be with specialized custom programming, anyone can use the toolset to make it happen. In fact, many programmers even think of the parallel processing engine as a type of “black box”. You send it data, it computes it however it does it, and on the far side comes the good, computed data." Full Story

Open source in grid computing: past, present and future

It is not long now until the first Open Source Grid and Cluster Conference, to be held in Oakland, California from 13-15 May 2008. This upcoming event has Grid luminary Ian Foster thinking about the role of open source in grid and cluster computing, in the past, present, and future.

"Overall, it’s an exciting time for open source grid and cluster computing. Much of this progress will be presented at the Open Source Grid and Cluster Conference. If you want to learn about what software exists, find out what people are doing with that software, and engage in discussions about what comes next, you should attend! If you do, I look forward to seeing you there." Full Story

OpenSolaris HPC Community

Many thanks to the InsideHPC blog for bringing us this story:

"The folks over at the OpenSolaris community have dedicated a community focus group to high performance computing. The community is designed to spark interest and exchange notes on utilizing the OpenSolaris platform for high performance computing. The community page lists four main purposes:

.: Provide a home for the OpenSolaris HPC distro

.: Highlight existing features and future enhancements relevant for high-performance computing

.: Provide a forum to discuss issues and possible solutions to problems faced by HPC developers

.: Provide pointers to pertinent resources and tools currently available.

The community page provides a list of participating development groups, including GridEngine, Xen, ZFS, pNFS, SAM-QFS, OpenMPI and Lustre. Thanks to Dan Templeton for raising this on the SGE mailing list." Full Story

Sun pushes its green IT agenda

Energy efficiency and utility computing are important to Sun in terms of sales due to the new European targets, but also because of the growing importance placed on environmental issues by customers, says Richard Barrington, UK head of sustainability and public policy.

"So if we're looking at doubling power consumption in the next 10 years in the UK, it means Sun has to deliver 40 percent more energy-efficiency savings. But if we provide the best environmental technology, it will be delivered by the truckload, and being able to drive energy efficiency is a good start," Barrington says. Full Story

Scientists Use Ranger to Explore Sub-Atomic Physics

HPC Wire brings us this story by TACC's Aaron Dubrow:

"Physics has led us deeper and deeper into the heart of matter in a quest to determine what the universe is made of. It's a real-world question that drives particle physicist Robert Sugar to dig in his heels and break out the big guns -- supercomputers like Ranger at the Texas Advanced Computing Center -- to help solve the numerical equations at the root of fundamental physics." Full Story

Allinea offers pathway to parallel

In this HPC Wire story, Allinea contends that, as the industry shifts to multicore architectures, "programming" and "parallel programming" must become synonymous. The company is working hard to advance the state of the art in debugging at scale as it collaborates with with TACC to investigate best practices for managing complexity on Ranger-scale problems:

"The processors we are seeing now, and will continue to see in at least the medium-term, will offer performance improvements only to those applications that can take advantage of many cores at one time. Since software customers generally expect applications to do more in less time, software developers have a strong incentive to parallelize their codes. But developers generally don't have the skills they need to make this change." Full Story

What cloud computing really means

InfoWorld ask the question: Is cloud computing anything more than just an updated version of utility computing?

"Cloud computing comes into focus only when you think about what IT always needs: a way to increase capacity or add capabilities on the fly without investing in new infrastructure, training new personnel, or licensing new software. Cloud computing encompasses any subscription-based or pay-per-use service that, in real time over the Internet, extends IT's existing capabilities." Full Story

Application Software Summit Takes on Multiphysics

A recent summit organized by the Council on Competitiveness brought together over 100 people from the HPC community to address the significant business and technical issues that impede multidisciplinary and multiphysics modeling.

"Increasingly companies must conduct multidisciplinary and full life-cycle product simulations to meet the competitive pressures of the global marketplace," explained Suzy Tichenor, vice president of the Council on Competitiveness. "But many companies that use high-performance computing cannot accomplish this task with current application software. This is really an area where the public and private sectors could cooperate to move forward." Full Story

Call for SC08 Analytics Challenge

It's time to start planning your submission for the SC08 Analytics Challenge. SC08 will be the fourth year of the Analytics Challenge, which highlights rigorous and sophisticated methods for solving complex, real-world problems. Notice of intent to participate in the Analytics Challenge is due May 1, 2008 and all Challenge submissions are due July 31, 2008.

"The Analytics Challenge will provide a forum for researchers and industry representatives to present complete solutions that embody all facets of high performance computing, such as comprehensive computational approaches, processing of large data sets and high-end visualization technology to display results. The Analytics Challenge will also bring forth interesting applications from a wide variety of areas. Past finalists have presented solutions in areas such as orthodontics, traffic control, atomic energy, geophysics and music classification. The finalists will be chosen by a panel of judges in August 2008 and will be invited to present their submissions at SC08. The overall winner will be announced at the SC08 Awards Ceremony." Flyer PDF

SLAC adds second Sun Modular Datacenter

To meet significant amounts of streaming data from ATLAS and GLAST simulation and mock data challenges, SLAC has installed their second Sun Modular Datacenter S20 (Project Blackbox). The largely self-contained data center is an efficient way to boost SLAC's overall computing capacity and is quicker and easier than building additional building space.

"We needed the additional capacity this year," Scientific Computing and Commuting Services (SCCS) Assistant Director Chuck Boeheim said. "With the first Blackbox working out very well, we invested in a second to meet increasing need." Full Story

Lustre to run on ZFS

According to this story from the 23rd annual High Performance Computer and Communications Conference, the Lustre development teams hopes to get a version of the ZFS-compatible Lustre released by the end of the year:

"Sun Microsystems developers are working on a version of the Lustre global file system that can run on the company's own next-generation Zetabyte File System (ZFS), said Peter Bojanic, director of Lustre engineering at Sun Microsystems." Full Story

For complete notes from the HPCC conference, check out this story on the Sun HPC Community Portal.

Parallel programming for the engineering masses

EE Times has a story on the mainstreaming of parallel programming tools:

"Until very recently, commercial high-level tools to support the development of technical computing applications for high-performance systems did not exist. Parallel programming was an esoteric art applied by specialists who focused on achieving maximum performance by using custom setups and low-level libraries and by tuning their applications for specific hardware." Full Story

Intel has designs on Petascale

The Register has a story about Intel's plans to get in the Petaflops race:

"We are aiming for the top of the top." Gelsinger declined to provide us with any specifics around an upcoming Intel-based monster. But you could see Intel working with some of the accelerator players to craft a "top" machine that teams co-processors with some upcoming "Nehalem" versions of Xeon. Or maybe Gelsinger is looking even further out at a system which taps the graphics capabilities of the upcoming Larrabee processor. Full Story

TACC Ranger: Computational Microscope

The power of the TACC Ranger supercomputer is enabling researchers to model the largest biomolecular apparatus to date:

"Schulten has spent his career extending the limits of microscopy by applying the immense power of supercomputers to molecular imagery. His “computational microscope” takes information from laboratory tests and turns it into dynamic, three-dimensional images with a powerful program Schulten created called NAMD (NAnoscale Molecular Dynamics, pronounced “NAM-dee”). Joining electron microscopy, x-ray crystallography, quantum chemistry and multi-scale molecular dynamics, with the massive parallel processing power of Ranger, the most powerful supercomputer in the world for open science research, Schulten’s molecular simulations are opening new realms of research that help us understand fundamental aspects of how life exists on earth.” Full Story

Sun Deploys Region's Biggest Grid at UAE

Sun Microsystems has deployed "one of the most advanced super grid computers in the world" at UAE University in the United Arab Emirates. The system is the region's biggest grid computing system.

Sun’s deployment consists of 100 blades, each with two quad core sockets, providing grid power of around 8.3 Tera Flops. The system is able to support both Intel-Xeon and AMD-Opetron flavors on the same type of blade server to maximize the usage of a wide range of HPC applications within one infrastructure. Users can log into the grid through Sun Thin Client technology – SunRay, that will be distributed among two separate laboratories each with 20 seats.

“Sun is delighted to have the UAE University as a new member to Sun’s High Performing Computing community- a community where it can engage with Sun engineers, partners and other customers in real time,” said Tarek Ayass, Education and Research Regional Manager for Sun Microsystems. “Sun’s open, scalable architecture of world-class components accelerates deployment, speeds results, and removes limitations. Our long-standing commitment to the high performance computing market delivers unmatched performance, flexibility, and ease of management. Sun’s goal is to provide support for educational and industrial research in the region, and collaborate with international Grid initiatives to lead international research toward the global grid; the next generation of the Internet,” Ayass added. Full Story

SAM-QFS Goes Open Source

On Tuesday, Sun added an important asset to our set of open source HPC software with the addition of SAM-QFS. Unlike other OpenSolaris projects, SAM and QFS are software products marketed and sold by Sun and its partners since 2001. The software has many commercial users in medical, media & entertainment, government, manufacturing, financial services, education that benefit from the services this software provides.

SAM is a heirarchical storage (HSM) and archive manager which provides automated policy management for users to easily administer large volumes of data stored across different tiers of storage from fast expensive disk to commodity storage devices to inexpensive, eco-friendly tape.

QFS is a Solaris-based cluster file system whose performance is ideally suited for data streaming and transactional applications. You can learn more at the SAM-QFS project site

So what does this mean to Sun? Project Manager Margaret Hamburger put it this way:

"First it shows that we're committed to being open. It's a strategy that affects they way we work, think about, and market technologies and products we offer. As a result, Sun is changing the storage marketplace. Unlike other vendors, Sun's storage software stack is open and with these major source contributions, Solaris gets even more interesting as a storage OS on which to build storage systems or appliances." Full Story

TACC Ranger explores the physics of flow

Professor P.K. Yeung of the Georgia Institute of Technology's School of Aerospace Engineering is using the TACC Ranger system to tackle the daunting physics of turbulence:

"Together with his research partners, Yeung recently began using Ranger, the world’s largest supercomputer for open-science research, at the Texas Advanced Computing Center (TACC) to resolve a turbulence grid of 68 billion (40963) grid points. The simulations will use up to 16,384 processor cores per simulation (or 131 trillion floating point operations per second), and this number will increase to 32,768 cores in the future (262 teraflops)." Full Story.

AMD to host Threadfest competition

AMD will host quarterly coding competitions throughout 2008 focused on advancing multi-threading application development for multi-core processor architectures:

"Called Multicore Threadfest, the competitions provide a forum for the software development community to demonstrate advanced multi-threading principles, techniques and best practices. Competition winners will be awarded cash prizes and earn recognition throughout the developer community, with each of four series prize winners having his or her code posted on the AMD Developer Central website (http://developer.amd.com), a resource portal for the software developer community." Full Story

One network, under IB

HPC Wire looks at Sun's proposal for a unified datacenter fabric based on InfinBand:

"Sun Microsystems seems to be getting impatient waiting for high performance Ethernet gear and FCoE standards, and has openly talked about going the InfiniBand route. In a recent EETimes article by Rick Merritt, Sun's John Fowler (executive VP of the Systems Division) said the company would be introducing a set of products this year unified on InfiniBand. While Fowler admitted they're continuing their involvement in the FCoE effort, he notes that Fibre Channel over InfiniBand should be available in 2008." Full Story

Photonic crystals for optical computing

European researchers have developed a relatively inexpensive way to make high-quality photonic crystals in the race to build all-optical chips:

"This is where photonic crystals come in. A combination of 3D photonic crystal optical switches and 2D waveguides could yield devices that are 10 or even 100 times smaller than those made at the moment. These could be used to assemble all-optical chips made entirely from silicon." Full Story

Intel to deliver six-core Xeon this year

Intel on Wednesday confirmed that it would ship the six-core Xeon processor in the second half of this year, putting to rest rumors about the processor's actual ship date.

Intel CEO Paul Otellini also described a new HPC platform called Larrabee during a speech at Intel's investor conference in Santa Clara, California, on Wednesday:

"Intel will make a real leap into the HPC segment with the Larrabee platform, which will have lots of cores, lots of threads and deliver performance "bar none," Otellini said. The platform will be delivered in the late-2009-to-2010 time frame, Otellini said." Full Story

For more information on Sun servers with Intel processors, check out the Sun/Intel alliance site.