NVIDIA Collaborates With Microsoft On High Performance GPU Computing
NVIDIAannounced it is working with Microsoft to promote NVIDIA Tesla GPUs forhigh performance parallel computing using the Windows HPC Server 2008operating system. “The coupling of GPUs and CPUs illustratesthe enormous power and opportunity of multicoreco-processing,” said Dan Reed, corporate vice president ofExtreme Computing at Microsoft. “NVIDIA’s work withMicrosoft and the Windows HPC Server platform is helping enablescientists and researchers in many fields achieve supercomputerperformance on diverse applications.
09/28/09
NVIDIA Releases Industry’s First Public OpenCL Conformant GPU Drivers and Performance Profiler
NVIDIAhas released the first public OpenCL conformant GPU drivers as well asa powerful performance profiling tool and an OpenCL Best PracticesGuide. The OpenCL Visual Profiler uses the extensive performanceinstrumentation in NVIDIA’s OpenCL drivers and hardwareperformance signals designed into NVIDIA GPUs to provide developerswith insight into performance bottlenecks and opportunities foroptimization. The OpenCL Best Practices Guide is designed to helpOpenCL developers programming for the CUDA architecture implement highperformance parallel algorithms and understand best practices for GPUComputing. The OpenCL drivers, Visual Profiler, and Best PracticesGuide are all available at:http://developer.nvidia.com/object/get-opencl.html.
09/28/09
NVIDIA Announces Live Webcast Coverage For GPU Technology Conference
NVIDIA’sinaugural GPU Technology Conference, which runs from Sept. 30 to Oct.2, has sold out. Keynote and general session addresses will be webcastlive. Webcast links, event coverage, and conference details can befound at www.nvidia.com/gtc.
09/28/09
CUDA Zone Hits New Milestone - Over 560 Apps and Papers
The560+ papers and apps posted on CUDA Zone cover a wide range ofscientific, professional, creative, and consumer applications. They aredeveloped / authored by scientists and engineers from all over theworld – nearly every major country is represented. Thespeedups they achieve over traditional methods are dramatic -- almostall of them have speedups in excess of an order of magnitude.
09/22/09
Real-Time Digital Holographic Microscopy Using the GPU – New on CUDA Zone
Indigital holographic microscopy (DHM), interfering wave-fronts from acoherent light-source are recorded on a sensor and the image digitallyreconstructed by a computer. The image yielded provides a quantitativemeasurement of the optical thickness of the specimen. In order toobtain a reconstructed image from a hologram, numerous calculations forthe Fresnel diffraction are required. The Fresnel diffraction can beaccelerated by the FFT (Fast Fourier Transform) algorithm. However,real-time reconstruction from a hologram is difficult. For example, ifone obtains a reconstructed image from a hologram whose size is 512×512using an Intel Core 2 Duo E6300 CPU, the calculation time for theFresnel diffraction takes about one second. In this paper, the authorsdescribe a real-time DHM system using CUDA to speed up thiscalculation. Authored by Tomoyoshi Shimobaba, Yoshikuni Sato, JunyaMiura, Mai Takenouchi, and Tomoyoshi Ito / Yamagata University, Japan.
09/22/09
Massively Parallel GPUs Accelerate Database Queries – New on CUDA Zone
Thispaper presents a new parallel indexing data structure that takes fulladvantage of the increasing thread-level parallelism emerging inmulti-core architectures. In this approach, the authors’ DataParallel Bin-based Index Strategy (DP-BIS) first bins the base data,and then partitions and stores the values in each bin as a separate,bin-based data cluster. In answering a query, the procedures forexamining the bin numbers and the bin-based data clusters offer themaximum possible level of concurrency; each record is evaluated by asingle thread and all threads are processed simultaneously in parallel.Submitted by Luke Gosink, Kesheng Wu, Wes Bethel, John D. Owens,Kenneth I. Joy / University of Calif. at Davis.
09/16/09
CUDA Drives Innovation in Broadcast and Film Production
IBC2009, AMSTERDAM—SEPT. 10, 2009—NVIDIA is making itpossible for production houses to work faster and easier with theimmense quantity of high-resolution data generated for HDTV, Blu-Rayand 4K digital cinema.
09/10/09
GPU Renderer for Maya - Furry Ball from Art and Animation Studio - New on CUDA Zone
NewGPU renderer features full Maya integration in Viewport, completerealtime dynamic fur and hair, bump mapping, shadows, reflection, andmore. 100-300 times faster on GPU as compared to CPU. Submitted by Artand Animation Studio, Czech Republic.
09/10/09
NVIDIA Releases Industry’s First OpenCL Performance Profiler for the GPU
NVIDIAReleases Industry’s First OpenCL Performance Profiler for theGPU New OpenCL Visual Profiler for Windows and Linux now available.Leveraging the extensive performance instrumentation inNVIDIA’s OpenCL drivers and hardware performance signalsdesigned into NVIDIA GPUs, the OpenCL Visual Profiler providesdevelopers with insight into performance bottlenecks and opportunitiesfor optimization.
09/09/09
CUDA Accelerates Digital Photo Library Management
SANTACLARA, Calif.—Sept. 8, 2009—NVIDIA announced that CyberLink MediaShow5, a new software program that organizes digital photos based on who isin them, is utilizing the CUDA parallel processing power of NVIDIAGeForce GPUs to search and sort photo libraries.
09/08/09
New Video on NVIDIA NEXUS for Visual Studio-based GPU Development
Checkout the new trailer video for NEXUS, NVIDIA's upcoming VisualStudio-integrated toolset for graphics and GPU Computing. NEXUSincludes powerful debugging and platform-wide performance tools. Moreinformation about NEXUS will be available at the GPU TechnologyConference, Sept. 30 - Oct. 2, 2009.
09/04/09
IEEE Spectrum: Why Graphics Processors Will Transform Database Processing
AuthorsAndrea di Blas and Tim Kaldewey, researchers at Oracle, write abouttheir work using graphics processors to comb through enterprisedatabases.
09/04/09
LAPACK for CUDA – New Linear Algebra Library from Magma Project
TheMAGMA project aims to develop a dense linear algebra library similar toLAPACK but for heterogeneous/hybrid architectures, starting withcurrent "Multicore+GPU" systems.
09/04/09
September CUDA Webinars. Sign up today!
Webinars on the basics of data parallel computing on GPUs leveraging NVIDIA’s CUDA architecture.
09/04/09
NVIDIA Delivers Comprehensive OpenCL Support under Snow Leopard
SANTACLARA, Calif. —Sep. 3, 2009— Apple’s new Snow Leopard operating system(OS) is the first OS to integrate OpenCL, a cross-platform openstandard that makes it possible for developers to tap into the vastgigaflops of computing power currently in the graphics processing unit(GPU) and use them for any application.
09/03/09
CUDA Supercharges Ray Tracing
Recentray tracing demos: 1) NVIDIA OptiX ray tracing engine, built on theCUDA architecture, enables developers to quickly createhighly-accelerated applications that employ ray tracing. 2) iray is anew rendering mode for the upcoming RealityServer and mental ray builton the CUDA architecture to deliver physically correct, globalillumination from scenes using real-world materials and lighting. 3)VRay – Chaos Group has independently implemented a CUDA-basedversion of their popular ray tracing renderer that shows +20X speedincrease over software version running on CPU.
08/28/09
NVIDIA CUDA Superhero Challenge Sparks Competitive Fire for GPU Computing Developers
SANTACLARA, Calif.—Aug. 27, 2009— NVIDIA Corporation today announced that itwill be working with TopCoder, a competitive software developmentcommunity, on the CUDA Superhero Challenge, a series of contests forcomputer programmers who will harness the parallel processing power ofthe NVIDIA® CUDA™ architecture to solve some of computing’s biggestchallenges.
08/27/09
NVIDIA GeForce GPUs and DirectCompute in Windows 7 Accelerate Digital Media Applications
SANTACLARA, Calif.—Aug. 21, 2009— By harnessing the parallel processingpower of NVIDIA® GeForce® GPUs and DirectCompute, a new technology inWindows 7, developers can take the power of graphics processors andmake it available for general-purpose computing and create cutting-edgedigital media applications.
08/21/09
Real-Time Fiber Tracking with CUDA – New on CUDA Zone
Fibertracking is a technique based on "diffusion tensor magnetic resonanceimaging" (DT-MRI) that allows a neurosurgeon to visualize the neuronalfibers in the brain. Submitted by Adiel Mittmann, Federal University ofSanta Catarina.
08/18/09
New CULA Linear Algebra Library From EM Photonics Brings GPU Computing To Millions Of Developers
SANTACLARA, Calif. —Aug. 17, 2009—EM Photonics today released a beta versionof CULA, an implementation of the industry-standard LAPACK linearalgebra library designed and optimized for NVIDIA’s massively parallelCUDA™-enabled graphics processing units (GPUs).
08/17/09
NVIDIA CUDA Technology Used To Recover Historic APOLLO 11 Man On The Moon Video
SANTACLARA, CA—AUGUST 5, 2009—When NASA’s originalApollo 11 moon landingvideo was accidentally destroyed, it seemed theworld had lost a visualrecord of one of man’s greatest achievements.
08/05/09
NVIDIA Releases Version 2.3 of the CUDA Toolkit
NVIDIAhas released a new version of the CUDA Toolkit and SDK for GPUComputing. The release includes performance improvements and expandedsupport for the cuda-gdb hardware debugger. Available for download.
07/22/09
NVIDIA Announces CUDA C Programming Best Practices Guide
NVIDIA’sfirst CUDA C Programming Best Practices Guide is designed to helpdevelopers programming for the CUDA architecture - using C with CUDAextensions - implement high performance parallel algorithms andunderstand best practices for GPU Computing.
07/09/09
PGI and NVIDIA Team to Deliver CUDA Fortran Compiler
ThePortland Group®, a wholly-owned subsidiary of STMicroelectronics andleading supplier of compilers for high-performance computing (HPC),today announced an agreement with NVIDIA under which the two companiesplan to develop new Fortran language support for CUDA GPUs.
06/23/09
Autodesk Leverages NVIDIA GPU Computing to Bolster Moldflow Software Performance
Autodesktoday announced it has significantly increased the performance of thelatest release of Autodesk Moldflow Insight 2010, part of its softwaresuite for plastics injection molding, by further leveragingcutting-edge GPU technology from NVIDIA
06/23/09
CUDA Toolkit and SDK 2.3 Betas now available for registered developers
Thisrelease includes support for all CUDA-capable GeForce, Quadro, andTesla products on WinXP/Vista/7, MacOS, and Linux. To join the GPUComputing Registered Developer Program, go to:http://developer.nvidia.com/page/registered_developer_program.html
06/23/09
NVIDIA and Supermicro Shatter 1U Server Performance Record
NVIDIACorporation and Supermicro announced the immediate availability of anew class of server that combines massively parallel NVIDIA® Tesla™GPUs with multi-core CPUs in a single 1U rack-mount server.
6/1/09
NVIDIA to Host Inaugural GPU Technology Conference
NVIDIA Corporation announced that its inaugural GPU Technology Conference will take place
September 30 to October 2, 2009. 5/26/09
NVIDIA Submits OpenCL 1.0 Driver to Khronos for Conformance Certification for Windows and Linux
NVIDIACorporation today announced its OpenCL™ 1.0 drivers for Windows XP andLINUX have been submitted to the Khronos OpenCL Working Group forcertification immediately after the conformance tests were approved.These pre-release drivers are now available to all NVIDIA GPU Computingregistered developers.
05/12/2009
NVIDIA CUDA TOOLKIT 2.2 RELEASED
NVIDIAannounced today it has released version 2.2 of the CUDA Toolkit and SDKfor GPU Computing. This latest release supports several significant newfeatures that deliver a major leap forward in getting the mostperformance out of NVIDIA’s massively parallel CUDA-enabled GPUs. Inaddition, version 2.2 of the CUDA Toolkit includes support for Windows7, the upcoming OS from Microsoft that embraces GPU Computing.
05/08/2009
GeoStar And NVIDIA To Transform Oil And Gas Industry In China
NVIDIACorporation and GeoStar, a leading Chinese geophysical servicesprovider, unveiled today the launch of a new hardware and softwaresolution that will transform seismic computation for oil and gascompanies in China.
04/29/2009
Cyberlink MediaShow Espresso Uses NVIDIA GPUs And CUDA Technology For Faster Video Encoding
NVIDIAtoday announced that Cyberlink has released its MediaShow Espresso, aneasy to use software application that converts digital video forplayback on different portable devices, such as the iPod, iPhone, orPSP at very high speeds with excellent quality.
04/28/2009
‘NERO MOVE IT’ Gets 5-Fold Performance Boost From NVIDIA CUDA Architecture
NVIDIAtoday announced that NERO has released an update to its Nero Move itsoftware that reduces video encoding time by up to five times byutilizing NVIDIA’s CUDA computing architecture built inside selectNVIDIA Graphics Processing Units (GPUs).
04/28/2009
NVIDIA Releases OpenCL™ Driver To Developers
NVIDIACorporation, the inventor of the GPU, today announced the release ofits OpenCL™ driver and software development kit (SDK) to developersparticipating in its OpenCL Early Access Program.
04/20/2009
Programming The CUDA Architecture: A Look At GPU Computing
GPUshave quickly surpassed CPUs in terms of computation speed. Nowprogrammers can use the CUDA architecture to help simplify theirimplementation.
04/09/2009
NVIDIA Introduces Industry's First Hardware Debugger and Profiler For GPU Computing
TheCUDA architecture continues to blaze a trail as the leading platformfor developing and running GPU Computing applications, with support forC, OpenCL™ , DirectX™ Compute, Fortran and other languages and APIs.The latest CUDA 2.2 Beta contains a host of significant new features.
04/08/2009
NVIDIA APEX Accelerates Physics Creation Pipeline
NVIDIAannounced NVIDIA APEX, a new capability for the PhysX SDK. APEXprovides artists, level designers, and game developers with easy-to-usetools that streamline the process of implementing scalable physicsacross multiple platforms.
03/25/2009
CSI-Style Video Enhancement For Consumers
MotionDSPreleased vReveal, an easy-to-use Windows application for PCs that fixescommon problems afflicting consumer-generated video. vReveal leveragesthe parallel processing power in NVIDIA CUDA-enabled GPUs.
03/24/2009
NVIDIA Looking For Next Great GPU Computing Company
NVIDIAlaunched the GPU Ventures Program, a new global initiative whose aim isto identify, support, and invest in early stage companies leveragingthe GPU for visual and other computing applications.
03/10/2009
Nero Achieves Video Acceleration with NVIDIA's CUDA
Nero,creators of liquid media technology, previewed the latest version ofNero Move it, which now features support for the NVIDIA CUDAarchitecture, at the CeBIT trade fair.
03/03/2009
New open-source software permits faster simulations of molecular motion on desktop computers
Whether vibrating in place or taking part in protein folding to ensure cells function properly, molecules are never still.
02/04/2009
NVIDIA Names Stanford's Bill Dally As Chief Scientist, VP of Research
BillDally, chairman of Stanford University's computer science department,joined NVIDIA as Chief Scientist and Vice President of NVIDIA Research.
01/28/2009
WIPRO To Offer CUDA Software Services To Global Customer Base
NVIDIA announced it is working closely with Wipro to provide CUDA™ professional services to their joint customers worldwide.
01/15/2009
NVIDIA CUDA Technology Dramatically Advances The Pace Of Scientific Research
Oncethought of as a technology used only for computer games, NVIDIA®GeForce® graphics processing units (GPUs) with CUDA™ technology are nowbeing used for the serious business of scientific computation.
12/17/2008
NVIDIA Adds OpenCL To Its Industry Leading GPU Computing Toolkit
NVIDIA announced its full support for the newly released OpenCL 1.0 specification from the Khronos Group.
12/09/2008
CUDA Cleans Up At SuperComputing Industry Awards
Everyyear at the Supercomputing (SC) convention(http://sc08.supercomputing.org), the organizing committee and mediapartners give awards for outstanding high performance computing (HPC)research and achievements.
12/08/2008
NVIDIA And CRAY To Deliver Tesla-Enabled CRAY CX1 Desk Side Super Computer
NVIDIACorporation (Nasdaq: NVDA) and Cray (Nasdaq GM: CRAY) today announcedthe availability of NVIDIA Tesla C1060 GPU Computing processors in thenew Cray CX1 line of supercomputers.
11/19/2008
DICE Puts Faith In NVIDIA PhysX Technology For Mirror's Edge
Inthe award-winning videogame Mirror’s Edge™, DICE, an Electronic ArtsInc. studio, introduces players to a new heroine named Faith.
11/19/2008
NVIDIA Tesla Gives Bull Customers A Revolutionary Performance Boost
Bull,a leading supplier of high performance computing (HPC) technologies, ispartnering with NVIDIA to provide the Tesla™ S1070 GPU Computing Systemas the accelerator option for their HPC solutions.
11/19/2008
NVIDIA And NEC Collaborate To Deliver GPU Computing Solutions To HPC Market
NVIDIAhas announced today that it has begun a close collaboration with NEC tointegrate NVIDIA® Tesla™ GPUs into its systems for the high performancecomputing (HPC) industry.
11/19/2008
NVIDIA Tesla TurboCharges High-Performance Computing Industry With HP Proliant Servers
NVIDIAtoday announced that the Tesla™ S1070 Computing System is now beingoffered in the highly successful range of HP ProLiant servers.
11/19/2008
NVIDIA Tesla Makes Personal SuperComputing A Reality
Today,scientific research is carried out on supercomputing clusters, a sharedresource that consumes hundreds of kilowatts of power and costsmillions of dollars to build and maintain.
11/18/2008
Mathematica Users Get 100x Performance Boost From NVIDIA CUDA
AtSC08, Wolfram Research will demonstrate a new version of Mathematica,the world’s most powerful general computational software, thatintegrates CUDA®, NVIDIA’s parallel GPU computing architecture.
11/18/2008
NVIDIA Demonstrates Powerful GPU Computing Solution From Lenovo At SC08
Today at SC08, NVIDIA demonstrated a Lenovo ThinkStation equipped with Tesla™ C1060 GPU Computing processing technology.
11/18/2008
Tokyo Tech Builds First Tesla GPU Based Heterogeneous Cluster To Reach Top 500
TheTokyo Institute of Technology (Tokyo Tech) today announced acollaboration with NVIDIA to use NVIDIA® Tesla™ GPUs to boost thecomputational horsepower of its TSUBAME supercomputer.
11/18/2008
OpenGeoSolutions Transforms Seismic Modeling With NVIDIA TESLA
Geophysicistsin the oil and gas industry are seeking more accurate images of whatlies beneath the earth. In order to find what’s been buried formillions of years, Calgary-based OpenGeoSolutions uses a techniquecalled “Spectral Decomposition” specifically to reveal geologicalinformation that goes beyond classic seismic resolution and detection.
11/18/2008
Nvidia’s $50 card destroys ATI’s $500 one or “Why ATI sucks in Folding?”
The Bright Side of IT
Asyou might already know, I am a bit enthusiastic when it comes todistributed computing. I’ve been looking for aliens through SETI@home,later with BOINC… but then, Folding@Home showed up and I became anenthusiast for this valuable project from Stanford University.
10/24/2008
Introduction, CPU and GPU differences
Digit-Life
Parallelcomputing has already entered the mass market and 3D games. Universaldevices with multi-core processors for parallel vector computing in 3Dgraphics reach high peak performance, CPUs cannot keep up with it.
10/21/2008
Is Your Personal Computer A CUDA-Enabled Speed Merchant?
Electronic Design
SometimesI don’t hear a rumble until it becomes a roar. I’m not sure if CUDA hasbecome a roar yet, but my ears have perked up based on a bunch ofannouncements I’ve received over the past few months. If CUDA hasn’tregistered on your radar yet, here’s a brief summary.
09/25/2008
GPUs Finding A New Role on Wall Street
Electronic Design
Oneof the new kids on Wall Street is GPU computing, a technology that ismaking inroads across nearly every type of HPC application. The vectorprocessing capabilites of GPUs makes them especially well-suited tofinancial analytics.
09/22/2008
GPUs Finding A New Role on Wall Street
HPCWire
Withclock speeds more or less stagnant now and the promise of multicore CPUscalability still a pipe dream, the data parallelism offered by GPUs isone way at least some applications can jump back on the performancecurve. The way Hanweck sees it, "from a technology standpoint, GPUs aregoing to change the way the world works."
09/22/2008
Nvidia Chip Speeds Up Imaging for Industrial Use
The New York Times
Energyexploration firms, clothing designers, medical companies and financialservices firms have also bought systems running on Nvidia chips. All ofthese companies share a common problem: they need hardware that cananalyze a vast quantity of data and do it much faster than standardcomputers.
09/22/2008
CUDA, Supercomputing for the Masses: Part 8
Dr.Dobb’s Portal
Optimizedlibraries often provide an easy way to improve performance ofapplications. When porting large legacy projects, libraries may be theonly real way to optimize for a new platform because code changes wouldrequire extensive validation efforts.
09/19/2008
NVIDIA to Offer Its New Chips in the New Cray Desktop
GigaOm
Aftermore than two years of pushing its scientific computing efforts,Nvidia’s graphics processors will be offered as an option in the newestline of Cray desktop supercomputers. The chipmaker plans to announcenext week that its Tesla chips can be used in the $25,000 Cray desktopsupercomputer, according to Nvidia spokesperson Andrew Humber. He saidNvidia has been in talks with Cray ever since the chipmaker announcedits Tesla line of graphics processors in 2007, but that this is thefirst deal the two companies have inked.
09/16/2008
NCSA to add 62 teraflops of compute power with new heterogeneous system
University of Illinois' National Center for Supercomputing Applications
Installationhas begun on a new computational resource at the National Center forSupercomputing Applications (NCSA) at the University of Illinois atUrbana-Champaign. Lincoln will deliver peak performance of 62.3teraflops and is designed to push the envelope in the use ofheterogeneous processors for scientific computing.
09/08/2008
Agilent Collaborates with NVIDIA for Simulation
Desktop Engineering Online
AgilentTechnologies Inc. announced its work with NVIDIA to accelerate signalintegrity simulations using NVIDIA's CUDA-based GPUs. The associationis expected to yield the commercial release of a GPU-enabled ADSTransient Convolution Simulator that will allow signal integritydesigners to run these simulations faster than before.
09/02/2008
CUDA, Supercomputing for the Masses: Part 7
Dr.Dobb’s Portal
CUDAand CUDA-enabled devices are co-evolving to deliver more performanceand capability with each new generation. NVIDIA's recent introductionof the GeForce 200-series and Tesla 10-series of products, shows therapidity of this evolution as roughly twice the hardware capability isnow available at the same price point of the previous line of productsplus the 200-series includes the addition of some valuable (andpotentially indispensable) new features.
08/20/2008
NVIDIA's GeForce GPUs Used for More Than Graphics
HPCwire
Newconsumer application pack uses NVIDIA CUDA technology to improveperformance beyond graphics on NVIDIA GeForce GPUs. Consumers wantblazing fast performance -- whether blasting their way through thelatest game or being socially responsible and sharing their PC'sprocessing power to help find cures for diseases.
08/12/2008
Larrabee, CUDA and the quest for the free lunch
TG daily
Opinion– Intel unveiled some key details about its upcoming Larrabeeaccelerator/discrete graphics architecture earlier this week, sparkingspeculation how this new technology will stack up to what is alreadyout there in the market.
08/06/2008
The Future Looks Bright for Teraflop Computing
Scientific Computing
Today'sGPGPUs are offering a large number of computational cores with a localon-card memory space. A teraflop of commodity computing capability bytoday's standards is outstanding! In the relatively near future, it islikely that we will see new products and many-core chips from multiplevendors. If your software can transition to utilize these newplatforms, the future is bright indeed.Happy teraflop computing!
August 2008
NVIDIA Recognizes University Of Utah As A Cuda Center Of Excellence
NVIDIA Corporation
NVIDIACorporation, the worldwide leader in visual computing technologies, andthe University of Utah today announced that the university has beenrecognized as a CUDA Center of Excellence, a milestone that marks thebeginning of a significant partnership between the two organizations.
07/31/2008
Processor Bifurcation
Linux Magazine
The processor market is diverging between two paths, the general and the predictable. Where does HPC hitch it’s wagon?
07/29/2008
CUDA, Supercomputing for the Masses: Part 6
Dr.Dobb’s Portal
Astutereaders of this series timed the two versions of the reverse arrayexample discussed in Part 4 and Part 5 and were puzzled about how theshared memory version is faster than the global memory version.
07/25/2008
Parallel computing with GPUs
InfoWorld
Writinghighly parallel code is hard, but many of us are going to need to learnto do it in the next few years, since computers are now getting morecores and bigger caches instead of faster clocks. Writing good parallelcode for symmetric multi-processor computers with shared memory is hardenough, but when it becomes asymmetric, more than a little art isrequired.
07/25/2008
NVIDIA Accelerates the Search for a Cure
HPCwire
StanfordUniversity's distributed computing program Folding@home has become amajor force in researching cures to life-threatening diseases such ascancer, cystic fibrosis, and Parkinson's disease by combining thecomputing horsepower of millions of processors to simulate proteinfolding.
07/24/2008
NVIDIA Keeps It Interesting
HPCwire
NVIDIAis continuing to push hard on CUDA, the company's C-based softwareenvironment for GPU computing. With last month's announcement of thefirst CUDA Center of Excellence at the University of Illinois atUrbana-Champaign, NVIDIA said it donated half a million dollars to theschool.
07/24/2008
Going to the Well
Advanced Imaging Pro
"3Dvisualization has revolutionized the understanding of seismic data,thanks to the performance provided by the GPU," said Jean BernardCazeaux, Vice President of the Visualization Sciences Group at MercuryComputer Systems. "GPUs allow much more than visualization; theyprovide amazing computing capabilities for interactive applications.Mercury has facilitated the interoperability of Open Inventor withNVIDIA's CUDA language, to provide application developers with aunique, integrated solution."
07/08/2008
Graphics Chips Help Supercomputers Become Commonplace
ComputerWeekly.com
Thesight of supercomputers in every home and office may soon become areality thanks to video games such as Grand Theft Auto. High-end 3Dgames need the fastest graphics chips to run well. This has drivengraphics cards makers to build ever-faster cards, and performance fromthe graphics processor on these cards is hundreds of times faster thanthe processor in a standard PC.
07/07/2008
Desktop Supercomputing
The Engineer
Applicationsthat have components of sequential and serial processing, such astranscoding digital video from one format to another divide the workbetween the CPU and the GPU to give about 20 times the performance ofjust using the CPU alone.
07/01/2008
CUDA, Supercomputing for the Masses: Part 5
Dr.Dobb’s Portal
Thelocal and global memory spaces are not cached which means each memoryaccess to global memory (or local memory) generates an explicit memoryaccess. So what does it cost to access (read or write, for example)each of the different memory types?
06/30/2008
Tesla 10 & CUDA 2.0: Technical Analysis & Performance - Page 1
Beyond 3D
CUDAwas announced along with G80 in November 2006, released as a publicbeta in February 2007, and then finally hit the Version 1.0 milestonein June 2007 along with the launch of the G80-based Tesla solutions forthe HPC market. Today, we look at the next stage in the CUDA/Teslajourney: GT200-based solutions, CUDA 2.0, and the overall state ofNVIDIA's HPC business.
06/26/08
GPGPUs Make Headway in Bioscience
HPCWire
We'realso exploring bioinformatics applications, but the really great thingabout the GPGPU and CUDA right now is that post-docs and universitiesare porting codes and putting them back into the public domain at anincredible rate. This means that the community effort can be used toleverage standard codes without a large investment. Everyone has a GPU,and CUDA can be gotten by just hitting the download button.
06/25/2008
More Details on Elemental's GPU Accelerated H.264 Encoder
AnandTech
Elemental'ssoftware, if it truly performs the way as seen here, has the potentialto be a disruptive force in both the GPU and CPU industries. On the GPUside it would give NVIDIA hardware a significant advantage over AMD'sGPUs, and on the CPU side it would upset the balance between NVIDIA andIntel. Video encoding has historically been an area where Intel's CPUshave done very well, but if the fastest video encoder ends up being aNVIDIA GPU -- it could mean that video encoding performance would bemicroprocessor agnostic, you'd just need a good NVIDIA GPU.
06/24/2008
Stanford releases beta Nvidia folding client
The Tech Report
“Atlast, Stanford University has released a beta version of the GPU2Folding@home client for Nvidia graphics cards. You can grab the clientfrom this post on the official FAH forums, although Stanford's AdamBeberg suggests users closely read the FAQ page to familiarizethemselves with the software first.”
06/20/2008
NVIDIA, CUDA and PhysX
EuroGamer
“3Dcard manufacturers shouldn't take this the wrong way, but it takes alot to make us crawl out of the communal Eurogamer bed (yes, all theEurogamer writers share a single large bed - we do it for frugality andcommunality, which remain our watchwords) and go to a hardwarepresentation. There's a nagging fear someone may talk maths at us andwe'd come home clutching the local equivalent of magic beans. And thenwe'll be laughed at by our fellow writers and made to sleep in thechilly end where the covers are thin and Tom left dubious stains.That's no fun at all.”
06/20/2008
Can you feel it?
Linux Magazine
Allof this sounds very familiar to the great cluster disruption. Take alook at the enabling factors list above. Like clusters, the cost to getin the game is minimal. There are over 70 million CUDA enabled GPUssitting in workstations out there. If you don’t have one, a the cost ofa basic GeForce video card was less than $100. As for the software, itis freely available. NVidia, quite wisely, makes the CUDA C compileravailable at no cost (and with no registration hassles). It isessentially the same cluster recipe, a low (or no) cost of entry, apossible big pay-off, and some spare time.
06/18/2008
NVIDIA's CUDA: The End of the CPU?
Tom's Hardware
CUDAis not a gimmick intended for researchers who want to cajole theiruniversity into buying them a GeForce. CUDA is genuinely usable by anyprogrammer who knows C, provided he or she is ready to make a smallinvestment of time and effort to adapt to this new programmingparadigm. That effort won’t be wasted provided your algorithms lendthemselves to parallelization.
06/18/2008
OptiTex to Use NVIDIA's CUDA Technology
TenLinks.com
"OptiTex'software is an ideal fit for NVIDIA as it leverages the combinedpersonalities of our CUDA enabled GPUs - rich graphics and dataintensive computation," said Andy Keane general manager of the GPUComputing business at NVIDIA. "OptiTex' software will deliver newlevels of creative freedom for designers."
06/16/2008
NVIDIA Looking to Take Computing to the Next Level
SFGate.com
“NVIDIAreleased a new set of GPUs that not only boast a crazy amount of speed,but come with the promise of helping take on a larger set of tasks bydelivering a lot more usable horsepower.”
06/16/2008
NVIDIA Releases 240-Core Graphics Processor
eWeek.com
“Tesla 10 series processor is Nvidia's latest offering for high-performance computing.”
06/16/2008
Nvidia and Stanford Finalizing Folding@Home Client for GeForce GPUs
tgdaily
“DuringNvidia Editor's Day, we learned that Nvidia and the Folding@Homeresearch group led by Vijay Pande are making final preparation tolaunch the first version of the Folding@Home client for Nvidia graphicsprocessors.”
06/13/2008
Apple Eyeing NVIDIA's CUDA Technology?
CNET
“One of the most important performance challenges facing CUDAApple's Worldwide Developers Conference is expected to cover theparallel tracks of Mac and iPhone software development, but the companymay have another aspect of parallelism to discuss.”
06/06/2008
CUDA, Supercomputing for the Masses: Part 4
Dr.Dobb’s Portal
“One of the most important performance challenges facing CUDA(short for "Compute Unified Device Architecture") developers is thebest use of local multiprocessor memory resources such as sharedmemory, constant memory, and registers.”
06/03/2008
NVIDIA Processor Has New Niche
Wall Street Journal
CUDAwas a breakthrough because it "makes things much, much easier forpeople developing applications because you no longer have to be agraphics expert," says Steve Briggs, vice president of systemsintegration at Headwave Inc., a Houston company that makes softwarethat crunches seismic data for the oil-and-gas industry.
03/26/2008
OpenCL is a trademark of Apple Inc. used under license to the Khronos Group Inc.
DirectX is a registered trademark of Microsoft Corporation.
|
