August 26, 2010
Here is a collection of highlights from this week's news stream as reported by HPCwire.
Numerical Algorithms Group Releases Fortran Builder 5.2
Spintronics Breakthrough Holds Promise for Next-Generation Computers
Bull Supercomputer Starts Work at Atomic Weapons Establishment
NICS to Add More Than 300 Teraflops to the NSF's Computing Capacity
Marshall University to Receive More than $525,000 to Enable Internet2 Access
AMD Announces Two New x86 Core Designs
LSU Research Group Part of DARPA Project to Create Advanced Computing Systems
Bielefeld University Selects ScaleMP for Molecular Physics Research
Platform Computing Introduces Low Cost Starter Pack for Private Cloud
AMD Appoints Donald Newell as Server Chief Technology Officer
NVIDIA Names Georgia Institute of Technology a CUDA Center of Excellence
Michael L. Norman Named Director of the San Diego Supercomputer Center
EM Photonics, U of Delaware Team to Develop Advanced Algorithms for Air Force
Shodor Attends TeraGrid '10, Wins Award
Supercomputing Software Ported to Windows in Belarus
Supercomputing, anytime, anywhere
Researchers from the Massachusetts Institute of Technology (MIT) and the Texas Advanced Computing Center (TACC) have come up with a pared down version of a supercomputing application for the Android smart phone. The team used TACC's Ranger supercomputer to analyze large, time-consuming simulations and create a "reduced model" that would work on a handheld device, specifically the Google Android smart phone. With the reduced model, the application becomes portable and can be used to generate real-time results. Basically any computation done on the smart phone can be thought of as a continuation of the work done on the supercomputer.
The release poses the question: "What if you could perform supercomputing calculations in real-time, on your smartphone, in any location?" An attention grabber, for sure. Let's just say that while an inexpensive handheld device will never compete with a multi-million dollar massive supercomputer on every level, with the right preparation and custom-crafted algorithms, results can indeed be comparable.
"Once you've created the reduced model, you can do all the computations on a phone," said David Knezevic, a post-doctoral associate in mechanical engineering at MIT, working in the lab of Prof. Anthony Patera.
Reduced models are not new, but the MIT/TACC team improved the error bounds, leading to more accurate and reliable results.
Knezevic, again: "We have a bound on how much accuracy we're losing with our reduced model, so we can say with rigor that we're doing supercomputing on a phone."
Here's more from the release:
The reduced model is constructed by focusing the supercomputer simulations on a range of parameters that are of interest to the user. Once the construction is finished, the model can be used to perform simulations for new parameters, nearly instantaneously, for use in real-world applications.
Using the smart phone application, researchers can change values, improve the error bounds by increasing the complexity of the local calculation, and even visualize the solution interactively in three dimensions.
Being able to run supercomputing-type apps on small, low-power devices like the smart phone is perfect for real-time field work. Possible uses being considered are landmine detection and other types of applied science and engineering, such as building design. The algorithm can also be used with linked devices acting as sensors that provide the data to the reduced order model. This could lead to apps such as real-time traffic reports, or "live" control systems for automobiles and aircraft for improved safety and performance. For more analysis on using a cluster of smart phones for real-time intelligent applications, check out this piece from Douglas Eadline.
Knezevic summed up the potential of the project thusly: "When you tell people you can solve a problem that would normally take two hours on Ranger in one second, with guaranteed error bounds, people instantly understand what model reduction is all about."
3PAR bidding war continues
Last week, Week in Review covered Dell's offer to buy 3PAR at $18 per share for a total of $1.15 billion cash. Exactly one week later, HP announced its counter offer: $24.00 per share, or $1.5 billion in cash.
Well, this morning it seemed that the bidding war was most likely over. 3PAR announced that it was accepting Dell's counter-counter-offer for $24.30 in cash per share, or almost $1.52 billion. The contract included a $72 million termination fee, sort of the reverse of a prenup, where 3PAR would have to pay Dell if it accepts any another offer.
Late in the day Thursday, HP showed that it is not ready to back down from this three-way tug of war, boosting its offer to $27 per share, about 1.69 billion. An amount that is over market price for the stock.
As of press time, Dell hasn't presented a counter-offer, but according to the original Dell/3PAR announcement, Dell has perpetual matching rights, meaning it has the right to match any counter-offer within three days.
The stock, by the way, is still trading above $26, more than double its price from before the takeover battle broke out.
The offer war between the three parties just proves how badly both Dell and HP want to be to fortify their storage offerings with 3PAR's technology.
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?