Archiwum dla Kwiecień, 2010

Nowa generacja kart Tesla C2050 (FERMI) już dostępna!!!

Napisany przez e-on w Aktualności dnia 28.04.2010

Przyjmujemy już zamówienia na dostawy kart TESLA C2050.

Dostępność kart serii C2070 jest przewidziane na wrzesień 2010r.

Nowa generacja architektury CUDA, karty NVIDIA Tesla™ C2050 i C2070, oznaczone nazwą kodową „Fermi”, to najbardziej zaawansowana architektura dedykowana obliczeniom GPU, jaka kiedykolwiek powstała. Dzięki ponad trzem miliardom tranzystorów oraz aż do 448 rdzeniom przetwarzania CUDA, zapewnia ona możliwości i wydajność superkomputerów opartych o 4 rdzeniowe układy CPU, przy 1/10 ceny oraz 1/20 zapotrzebowania na energię tradycyjnych serwerów korzystających wyłącznie z układów CPU.

Tesla C2050 Board 3GB ECC GDDR5, do 0.52 TFlops (Double Precision)

Sugerowana cena netto: 9′500,00 PLN

Specyfikacja techniczna:

Form Factor 9.75″ PCIe x16 form factor
# of Tesla GPUs 1
# of CUDA Core 448
Frequency of CUDA Cores 1.15 GHz
Double Precision performance 515 Gflops
Single Precision performance 1.03 Tflops
Total Dedicated Memory*

Tesla C2050
 3GB GDDR5
Memory Speed 1.55 GHz
Memory Interface 384-bit
Memory Bandwidth 148 GB/sec
Power Consumption 225W TDP
System Interface PCIe x16 Gen2
Thermal Solution Active Fansink
Software Development Tools CUDA C/C++/Fortran, OpenCL, DirectCompute Toolkits. NVIDIA Parallel Nsight™ for Visual Studio

Brak komentarzy

Konferencja GPU Technology Conference (GTC), 20-23 września 2010, San Jose, USA

Napisany przez e-on w Aktualności dnia 28.04.2010

W dniach 20-23 września w Kalifornijskim San Jose organizowana jest konferencja GPU Technology Conference (GTC).

W chwili obecnej przyjmowane są zgłoszenia prezentacji badań oraz prowadzenia sesji ze strony klientów branżowych oraz środowiska akademickiego.

W przypadku pytań dotyczących składania wniosków – kontakt w języku angielskim an adres: vcrimmins@nvidia.com
Więcej o konferencji można znaleźć na stronie: http://www.nvidia.com/object/gpu_technology_conference
Zgłoszenie do listy mailingowej związanej z konferencją znajduje się na stronie: http://www.nvidia.com/object/email_updates

Brak komentarzy

Tygodniowy newsletter “CUDA: Week in Review.” – 23.04.2010

Napisany przez e-on w Aktualności dnia 28.04.2010

CUDA: Week in Review

Friday, April 23, 2010, Issue #18
WELCOME
Reminder: The GPU Technology Conference (GTC), Sept. 20-23, is now accepting proposals for sessions and research posters from industry and academia.
– For questions about submissions, email: vcrimmins@nvidia.com
– Learn more here: http://www.nvidia.com/object/gpu_technology_conference
– Join the email update list here: http://www.nvidia.com/object/email_updates
CUDA NEWS
Books! Books! Books!
GPU computing is being taught at over 330 universities around the world, and now there are at least 10 educational books available/in progress on the topic of programming GPUs with CUDA C and OpenCL. This includes the brand-new “Basics of CUDA Technology” by A.V. Boreskov of Russia. The books reflect the international momentum behind the GPU, as 4 are in Japanese, 3 in Chinese, 2 in English, and 1 in Russian. See them listed here: http://www.nvidia.com/object/cuda_books.html.
New Update from AccelerEyes
AccelerEyes released a major update to its Jacket software platform for MATLAB (Jacket is a productivity platform for GPU acceleration). Jacket version 1.3 incorporates many new features, including support for double precision linear algebra using the CUDA architecture. Linear algebra has applications in the natural sciences (e.g. chemistry, physics, astronomy) as well as the social sciences (e.g. economics). “Initial results are very promising in Jacket 1.3, with an 8-10 times performance increase,” said Norman White, Faculty Director at NYU’s Stern Center for Research Computing. For more details, see: http://www.accelereyes.com/resources/announce and http://www.accelereyes.com/news/version13
CUDA ZONE
New on CUDA Zone: N-body Code on GPUs
Extract: “We present the characteristics and performance, both in terms of computational speed and precision, of a numerical code which integrates the equation of motions of N ‘particles’ interacting via Newtonian gravitation and moving in an external galactic smooth field. The code, NBSymple, has been parallelized twice by means of the CUDA architecture to make the evaluation as fast as possible on NVIDIA TESLA C1060s. The code works in single precision floating point arithmetic or in double precision.” Authors: R. Capuzzo-Dolcetta, A. Mastrobuono-Battisti, D. Maschietti (Dept. of Physics, Sapienza, Univ. of Rome, Italy). [Ed. note: An N-body simulation is a simulation of a dynamic system of particles, usually under the influence of physical forces, such as gravity.] See: http://is.gd/bCOmG
CUDA Zone Submissions
Have a CUDA-related paper, research, or app? Show it on CUDA Zone: http://is.gd/8G3E4
CUDA JOBS
The Irish Centre for High-End Computing (ICHEC) will provide funding for a select number of talented undergraduate students to engage in 10-week summer projects in Ireland. The projects are related to high-performance computing and computational science and will be under the supervision of ICHEC scientists and systems staff. See: http://is.gd/bBgc3
CUDA EVENTS
International Supercomputing Conference
ISC will be held in Hamburg, Germany, May 30-June 3. NVIDIA and its partners will be on hand to present HPC solutions, tutorials, and sessions. See: http://www.nvidia.com/object/isc2010.html
CUDA EDUCATION
NEW: SagivTech CUDA Training
May 10-12, Ra´anana, Israel. SagivTech offers CUDA courses in Israel and Europe, including customized training for organizations. Expertise in state-of-the-art image processing. See: http://www.sagivtech.com/24054.html
GPU Computing Webinars (CUDA C, OpenCL, Parallel Nsight and more…)
See schedule: http://developer.nvidia.com/object/gpu_computing_online.html
Acceleware-Certified CUDA Training
May 19-20, Silicon Valley: http://www.acceleware.com/index.cfm/cuda-training/may19sunnyvale/
GPGPU Conferences and Symposia
– GPU Computing in the Oil & Gas Industry (Microsoft/NVIDIA), May 12, Houston:
https://msevents.microsoft.com/CUI/EventDetail.aspx?EventID=1032446248&culture=en-US
– Parallel Execution of Sequential Programs on Multi-Core Architectures. June 20,
France: http://cccp.eecs.umich.edu/pespma/cfp.html
– GPUs in Chemistry and Materials Science, June 28-30, Univ. of Pittsburgh:
http://www.sam.pitt.edu/education/gpu2010.register.php
– Parallel Symbolic Computation 2010 (PASCO), July 21-23, France:
http://pasco2010.imag.fr/contest.html
– Symposium on Chemical Computations on GPGPUs, Aug. 22-26, Boston:
http://illinois.edu/lb/article/2101/36281
– UnConventional High Performance Computing 2010 (UCHPC 2010), Aug. 31-Sept. 1, Italy:
http://www.lrr.in.tum.de/~weidendo/uchpc10/
– GPU Technology Conference 2010, Sept. 20-23, San Jose, Calif.:
http://www.nvidia.com/gtc (now accepting proposals from industry and academia)
CUDA and GPU Computing Courses
Over 330 universities are teaching CUDA and GPU Computing courses.
– See the list: http://www.nvidia.com/object/cuda_courses_and_map.html
CUDA RESOURCES
CULA LAPACK
GPU-accelerated linear algebra library (from EM Photonics): http://www.culatools.com
(See recent CULA blog post titled “Initial Fermi Performance”: http://www.culatools.com/blog/2010/04/16/11-initial-fermi-performance)
NVIDIA Parallel Nsight
Download the Parallel Nsight Beta: www.nvidia.com/nsight
CUDA Toolkit
Download CUDA Toolkit 3.0: http://bit.ly/aKCENp
CUDA Documentation
Download developer guides and documentation: http://developer.nvidia.com/object/gpucomputing.html
CUDA Books
Programming Massively Parallel Processors by D. Kirk, W. Hwu: http://is.gd/7bNYP
CUDA ON THE WEB
– Follow CUDA & GPU Computing on Twitter: www.twitter.com/gpucomputing
– Network with other developers: www.gpucomputing.net
– Stayed tuned to GPGPU news and events: www.gpgpu.org
– Learn more about CUDA on CUDA Zone: www.nvidia.com/cuda
– CUDA on YouTube: http://www.youtube.com/nvidiacuda
About CUDA
CUDA is NVIDIA’s parallel computing hardware architecture. NVIDIA provides a complete toolkit for programming on the CUDA architecture, supporting standard computing languages such as C, C++, and Fortran as well as APIs such as OpenCL and DirectCompute.

See previous issues of CUDA: Week in Review: http://www.nvidia.com/object/cuda_week_in_review_newsletter.html
Copyright © 2010 NVIDIA Corporation. All rights reserved. 2701 San Tomas Expressway, Santa Clara, CA 95050.

Brak komentarzy

Tygodniowy newsletter “CUDA: Week in Review.” – 09.04.2010

Napisany przez e-on w Aktualności dnia 14.04.2010

CUDA: Week in Review

Friday, April 9, 2010, Issue #16
WELCOME
CUDA NEWS
Italian Supercomputing Center Improves Flexibility, Usability
CASPUR, a computing consortium in Rome, manages a high-powered processing center open to the Italian national scientific community. Researchers in biology, chemistry, medicine, mathematics, and materials science use the center’s high performance computing (HPC) clusters to run simulations and process large amounts of data. To improve the usability of the center, CASPUR adopted a pilot cluster that utilizes NVIDIA Tesla GPUs and CUDA C running on the Windows HPC Server 2008 operating system. In tests, CASPUR found that the cluster’s performance exceeded that of previous systems. “We noted performance from 10 to more than 100 times greater than that of single-processor systems,” comments Nico Sanna, Senior Technology Manager, CASPUR. See Microsoft case study: http://is.gd/b794h
Tokyo Institute of Technology Selected as a CUDA Center of Excellence
Tokyo Institute of Technology (Tokyo Tech) has been named Japan’s first CUDA Center of Excellence in recognition of its pioneering activities in parallel computing. Tokyo Tech is the 10th CUDA Center of Excellence, joining other institutions including Cambridge University, Chinese Academy of Sciences, Harvard University, National Taiwan University, Tsinghua University, University of Illinois at Urbana-Champaign, University of Maryland, University of Tennessee, and University of Utah. Tokyo Tech GSIC (Global Scientific and Information Computing Center) was the first supercomputing center to achieve a Top 500 ranking using GPUs. The Tokyo Tech TSUBAME 1.2 supercomputer uses 170 NVIDIA Tesla S1070 computing systems. For more info, see: http://cudacoe.m.gsic.titech.ac.jp/
CUDA ZONE
New on CUDA Zone: GPU-based Framework for Simulating Cortically-Organized Networks
Extract: “Computational models whose organization is inspired by the cortex are increasing in both number and popularity…. These models present two practical challenges. First, they are computationally intensive. Second, while the operations performed by individual cells are typically simple, the code needed to keep track of network connectivity can quickly become complicated…. We have created a programming framework called CNS (Cortical Network Simulator). CNS models are automatically compiled and run on a GPU, typically 80-100X faster than on a single CPU…. Authors: J. Mutch, U. Knoblich, T. Poggio, Center for Biological & Computational Learning, McGovern Institute for Brain Research, MIT.

Background on MIT’s Center for Biological & Computational Learning: “CBCL was founded with the belief that learning is at the very core of the problem of intelligence, both biological and artificial, and is the gateway to understanding how the human brain works and to making intelligent machines. CBCL studies the problem of learning within a multidisciplinary approach. Its main goal is to nurture serious research on the mathematics, the engineering, and the neuroscience of learning.” See: http://cbcl.mit.edu/ and http://is.gd/bjmOv
CUDA Zone Submissions
Have a CUDA-related paper, research, or application? Show it on CUDA Zone: http://is.gd/8G3E4
CUDA JOBS
Walt Disney Animation Studios seeks a senior development software engineer to join the team at the Burbank, Calif. studio. Responsibilities include: Re-engineer applications and libraries to leverage parallelism of emerging massive multicores; Be an evangelist for multicore computing throughout the organization. Requirements include: Practical experience with C++ or C; Experience with programming on the GPU using OpenGL, OpenCL, CUDA; Strong mathematical fundamentals, including linear algebra and numerical methods.
– See posting here: http://is.gd/bjmqm
– See more CUDA/GPU computing jobs here: http://is.gd/bjmJT
CUDA Education
GTC 2010: Call for Submissions Now Open
The GPU Technology Conference (GTC) 2010 will be held Sept. 20-23, 2010 in San Jose, Calif. Developers, researchers, scientists, and entrepreneurs are invited to submit proposals on topics related to the burgeoning GPU ecosystem. See: http://www.nvidia.com/object/call_for_submissions.html
Industry Event: Visual Studio Launch Conference
On April 12, Microsoft will launch Visual Studio 2010 and Silverlight 4 at the Microsoft Visual Studio Launch Conference, Las Vegas. NVIDIA will be on hand to demonstrate NVIDIA Parallel Nsight. See: http://www.devconnections.com/shows/SP2010VS
Seminar: GPU Computing in the Oil & Gas Industry
On May 12, Microsoft and NVIDIA will team up to offer a seminar/workshop for application programmers in the oil & gas. To register, visit: https://msevents.microsoft.com/CUI/EventDetail.aspx?EventID=1032446248&culture=en-US
GPGPU Conferences and Symposia
– Symposium on Applications of GPUs in Chemistry and Materials Science, June 28-30,
University of Pittsburgh: http://www.sam.pitt.edu/education/gpu2010.register.php
– Parallel Symbolic Computation 2010 (PASCO), July 21-23, France:
http://pasco2010.imag.fr/contest.html
– Symposium on Chemical Computations on GPGPUs, Aug. 22-26, Boston:
http://illinois.edu/lb/article/2101/36281
– Workshop on UnConventional High Performance Computing 2010 (UCHPC 2010),
Aug. 31-Sept. 1, Naples, Italy (with Euro-Par 2010):
http://www.lrr.in.tum.de/~weidendo/uchpc10/
Acceleware-Certified CUDA Training
Calgary, April 12-16: http://www.acceleware.com/index.cfm/cuda-training/apr12calgary/
Silicon Valley, May 19-20: http://www.acceleware.com/index.cfm/cuda-training/may19sunnyvale/
GPU Computing Webinars (CUDA C, OpenCL, Parallel Nsight and more…)
See schedule: http://developer.nvidia.com/object/gpu_computing_online.html
CUDA and GPU Computing Courses
Over 320 universities are teaching CUDA and GPU Computing courses. See the list: http://www.nvidia.com/object/cuda_courses_and_map.html
CUDA Books
“Programming Massively Parallel Processors,” by D. Kirk and W. Hwu. Available on Amazon.com: http://is.gd/7bNYP
CUDA DOWNLOADS
– NEW: NVIDIA Performance Primitives (NPP) library now available for CUDA Toolkit 3.0:
http://is.gd/bjpUT
– Download CUDA Toolkit 3.0: http://bit.ly/aKCENp
– Download Developer Guides: http://developer.nvidia.com/object/gpucomputing.html
CUDA ON THE WEB
– Follow CUDA & GPU Computing on Twitter: www.twitter.com/gpucomputing
– Network with other developers: www.gpucomputing.net
– Stayed tuned to GPGPU news and events: www.gpgpu.org
– Learn more about CUDA on CUDA Zone: www.nvidia.com/cuda
– CUDA on YouTube: http://www.youtube.com/nvidiacuda
About CUDA
CUDA is NVIDIA’s parallel computing hardware architecture. NVIDIA provides a complete toolkit for programming on the CUDA architecture, supporting standard computing languages such as C, C++, and Fortran as well as APIs such as OpenCL and DirectCompute.

See previous issues of CUDA: Week in Review: http://www.nvidia.com/object/cuda_week_in_review_newsletter.html
Copyright © 2010 NVIDIA Corporation. All rights reserved. 2701 San Tomas Expressway, Santa Clara, CA 95050.

Brak komentarzy

Tygodniowy newsletter “CUDA: Week in Review.” – 02.04.2010

Napisany przez e-on w Aktualności dnia 07.04.2010

CUDA: Week in Review

Friday, April 2, 2010, Issue #15
WELCOME
CUDA NEWS
Advancing the Development of RF Systems
People rely on radio frequency (RF) systems in all kinds of everyday devices, from cell phones and TVs to remote controls and set-top boxes. RF systems are also used in medical devices for the treatment of health problems such as rapid heartbeat, joint pain, and sleep apnea. Researchers at Aalborg University in Denmark are applying GPUs and the AccelerEyes Jacket platform for MATLAB to advance the theory, design, and implementation of RF systems. A team led by Dr. Torben Larsen is using CUDA and AccelerEyes to develop new algorithms that enable RF systems to run in a fraction of the time that it takes on a CPU. Data interpolation and Fast Fourier Transforms (FFT), for example, are common tasks performed in signal processing applications. Dr. Larsen’s team found that these tasks run 25 to 35 times faster with NVIDIA GPUs and Jacket as compared to CPUs. By increasing the speed at which RF systems can be developed and tested, the team aims to help bring effective and safe RF-enabled devices to market more quickly. Read more here: http://is.gd/b8Q4l
CUDA APPS
Faster Machine Learning Techniques Using GPUs
Support Vector Machine (SVM) is one of the most commonly applied machine learning techniques for classification. It’s used across areas such as web indexing and classification, text classification, image recognition, bio-informatics, predictive financial models, and business analytics. Two new CUDA GPU-enabled SVM software packages are now available as potential drop-in replacements for libSVM (the most popular SVM implementation) showing speedups of 10 to 70X over libSVM. For info on these packages, see:
– MultiSVM Multiclass SVM: http://code.google.com/p/multisvm
– cuSVM (includes MATLAB MEX wrapper): http://patternsonascreen.net/cuSVM.html
CUDA ZONE
New on CUDA Zone: Allinea Announces Tools for CUDA Architecture
U.K.-based Allinea Software announced a pre-release version of the Distributed Debugging Tool (DDT) for the CUDA architecture, coinciding with the release of CUDA Toolkit 3.0. Over the past year, Allinea has collaborated with the French Commissariat a l’Energie Atomique (CEA) to develop CUDA-specific features within its DDT product. Demonstrations of DDT for CUDA will take place at the International Supercomputing Conference in Hamburg, Germany, May 30-June 5. David Lecomber, CTO of Allinea Software, comments: “We are delighted to see the results of our collaboration with the CEA making it into our mainstream DDT product. We are now able to provide application developers with a single tool that can debug hybrid MPI, OpenMP, and CUDA applications on a single workstation or GPU cluster.” See: http://is.gd/b7bh5
CUDA Zone Submissions
Have a CUDA-related paper, research, or application? Show it on CUDA Zone: http://is.gd/8G3E4
CUDA JOBS
CST – Computer Simulation Technology AG in Darmstadt, Germany develops and markets high performance software for the simulation of electromagnetic fields in all frequency bands. Its customers represent a range of industries, from telecommunications and automotive to electronics and medical. The company is seeking a full-time CUDA developer for development of 3D electromagnetic simulation solutions.
– See posting here: http://is.gd/b75ZH
– See more CUDA/GPU computing jobs here: http://is.gd/91IEu
CUDA Education
Seminar: GPU Computing in the Oil & Gas Industry
GPU computing is rapidly being adopted in the oil & gas industry, especially for seismic processing and reservoir simulation. On May 12, Microsoft and NVIDIA will team up to offer a seminar/workshop for application programmers in this field. The event will provide an overview of NVIDIA Parallel Nsight for Microsoft Visual Studio, including a look at how this development environment can be used for debugging source code transparently on GPU hardware and profiling applications. To register, visit: https://msevents.microsoft.com/CUI/EventDetail.aspx?EventID=1032446248&culture=en-US
GPGPU Conferences and Symposia
– Parallel Symbolic Computation 2010 (PASCO), July 21-23, France. Enter the Computer
Algebra Parallel Programming Contest. See: http://pasco2010.imag.fr/contest.html
– Workshop on UnConventional High Performance Computing 2010 (UCHPC 2010),
Aug. 31-Sept. 1, Naples, Italy (with Euro-Par 2010). Submission deadline: June 14.
See: http://www.lrr.in.tum.de/~weidendo/uchpc10/
– Symposium on Chemical Computations on GPGPUs, Aug. 22-26, Boston. Abstracts due
April 5. Best paper wins an NVIDIA Tesla C2050. See: http://illinois.edu/lb/article/2101/36281
– Symposium on Applications of GPUs in Chemistry and Materials Science, June 28-30,
University of Pittsburgh. See: http://www.sam.pitt.edu/education/gpu2010.register.php
Acceleware-Certified CUDA Training
Silicon Valley, May 19-20: http://www.acceleware.com/index.cfm/cuda-training/may19sunnyvale/
Calgary, April 12-16: http://www.acceleware.com/index.cfm/cuda-training/apr12calgary/
GPU Computing Webinars (CUDA C and OpenCL)
See schedule: http://developer.nvidia.com/object/gpu_computing_online.html
CUDA and GPU Computing Courses
Over 320 universities are teaching CUDA and GPU Computing courses. See the list: http://www.nvidia.com/object/cuda_courses_and_map.html
Industry Event: Visual Studio Launch Conference
On April 12, Microsoft will launch Visual Studio 2010 and Silverlight 4 at the Microsoft Visual Studio Launch Conference, Las Vegas. NVIDIA will be on hand to demonstrate NVIDIA Parallel Nsight. See: http://www.devconnections.com/shows/SP2010VS
CUDA DOWNLOADS
– Download CUDA Toolkit 3.0: http://bit.ly/aKCENp
– Download Developer Guides: http://developer.nvidia.com/object/gpucomputing.html
CUDA RESOURCES
– GPU Technology Conference (GTC) 2010 – Sept. 20-23, San Jose, Calif: www.nvidia.com/gtc
– GPGPU Book: Programming Massively Parallel Processors, by D. Kirk, W. Hwu:
http://is.gd/7bNYP
CUDA ON THE WEB
– Follow CUDA & GPU Computing on Twitter: www.twitter.com/gpucomputing
– Network with other developers: www.gpucomputing.net
– Stayed tuned to GPGPU news and events: www.gpgpu.org
– Learn more about CUDA on CUDA Zone: www.nvidia.com/cuda
– CUDA on YouTube: http://www.youtube.com/nvidiacuda
About CUDA
CUDA is NVIDIA’s parallel computing hardware architecture. NVIDIA provides a complete toolkit for programming on the CUDA architecture, supporting standard computing languages such as C, C++, and Fortran as well as APIs such as OpenCL and DirectCompute.
Copyright © 2010 NVIDIA Corporation. All rights reserved. 2701 San Tomas Expressway, Santa Clara, CA 95050.

Brak komentarzy

Znamy ceny i dostępność kart rodziny FERMI.

Napisany przez e-on w Aktualności dnia 01.04.2010

Poniżej orientacyjne ceny kart w nowej architekturze FERMI dostępnych od połowy 2010 roku:

Tesla C2050 Board 3GB ECC GDDR5, do 0.52 TFlops (Double Precision)

Sugerowana cena netto: 9’500,00 PLN

Tesla C2070 Board 6GB ECC GDDR5, do 0.63 TFlops (Double Precision)

Sugerowana cena netto: 14’900,00 PLN

Brak komentarzy