CUDA:WeekinReview

Today’s topics include:

·CUDA Spotlight on Dr. Anders Eklund, Virginia Tech, who usesCUDA-accelerated fMRI (functional magnetic resonance imaging) tostudy the humanbrain.

·Researchers at the Universities of Sheffield andSussex hope to create anaccurate computer simulation of a honey bee brain usingGPUs.

·A thesis project at San Diego State uses CUDA toaccelerate animal noseprintidentification.

Feel free to forward to partners andcolleagues.

Mon., Oct. 8,2012, Issue #81

Read Newsletter Online | See Previous Issues

Welcome toCUDA: WEEK INREVIEW, a news summary for the worldwide CUDA,GPGPU and parallel programming community.

CUDA TECH TIP: How to detect CUDA compute capability at compiletime. See tip below.

YOUR CUDA STORY ATSC12: Let us know howyou use CUDA in 140 characters or less. Your submission will beconsidered for display on NVIDIA’s SC12 booth (#2217) and website.
Reply to the one-minute survey.

CUDA SPOTLIGHT

Using GPUs to Study the HumanBrain
This week’s Spotlight is on Dr. Anders Eklund, a postdoc atthe Virginia Tech Carilion Research Institute. Anders was a CUDASpotlight in 2011 while a student at Linkoping University inSweden. We caught up with him recently to learn about his currentwork. Read the interview here.

CUDA NEWS

Robotic Bees?
The i09 website reports on a GPU computing project at theUniversities of Sheffield and Sussex, where scientists "are hopingto create the first accurate computer simulation of a honey beebrain - and then upload it into an autonomous flying robot…. Theresearchers hope a robotic insect could supplement or replace theshrinking population of honey bees." See: http://bit.ly/QGzkAH

Chemistry and LifeScience
Equip@meso will hold its first-ever scientific event - "Chemistryand Life Science: Numerical Simulation to HPC" - on Oct. 18 at theUniversity of Strasbourg. Equip@meso is a France-based initiativeto create a network of supercomputers for calculation-intensivetasks. See: http://hpc.unistra.fr

CUDA:WeekinReview
F# on theGPU
At the "F#unctional Londoners" Meetup on Oct. 18, attendees willhear a talk by Daniel Egloff on F# on the GPU with Alea.CUDA. Dr.Egloff is the founder of QuantAlea. See: http://www.meetup.com/FSharpLondon/events/83105582/

HPC SolutionsWorkshop
Dell is holding an event called "Enabling Discovery with Dell HPCSolutions" on Oct. 25 in Baltimore, Maryland. The workshop istargeted to researchers and developers. Speakers include:
•TamasBudavari, Johns Hopkins University
•AnupMahurkar, University of Maryland Baltimore
•JeffreyB. Klauda, University of Maryland College Park
•KevinHildebrand, University of Maryland College Park
•JonathanBentz, NVIDIA
See: http://bit.ly/PG8apz

CUDAConsulting
Training, programming, and project development services areavailable from CUDA consultants around the world. To be consideredfor inclusion on list, email: cuda_week_in_review@nvidia.com (with CUDA Consultingin subject line).

GPU Computing onTwitter
For daily updates about GPU computing and parallel programming,follow @gpucomputing on Twitter.

GPU THESIS WATCH

Title: CUDAImplementation of Parallel Algorithms for Animal NoseprintIdentification
Author: Vincent Stanley Dayes, San Diego State University
Advisor: William Root, San Diego State University
URL: http://sdsu-dspace.calstate.edu/bitstream/handle/10211.10/2123/Dayes_Vincent.pdf

CUDA JOB OF THE WEEK

Ventana MedicalSystems, a member of theRoche Group, is seeking a Senior Imaging Scientist withexpertise in image processing and computer vision to work with acollaborative team of engineers and scientists on initiativespertaining to digital pathology. Familiarity with CUDA a plus. See:http://bit.ly/Ue6N9c

NEW ON THE NVIDIA BLOG

back to the top

Stay up to dateby reading the NVIDIA blog:
GPU-Accelerated Computing Surges in Russia, bySumit Gupta
Synerscope: Data Analysis for the Rest of Us,by Brian Caulfield

GPU MEETUPS

back to the top

Find a GPU Meetupin your location, or start one up. Upcoming meetings include:
Silicon Valley, Oct. 8
Paris, Oct. 18 (special full day event)
Brisbane, Oct. 25
New York, Oct. 30

Note: If you would be interested joining a GPU Meetup inMinneapolis, Minn., see: http://linkd.in/Vip04N

CUDA TECH TIP

back to the top

Detect CUDA Compute Capability at CompileTime
When nvcc compiles a .cu file, it defines the __CUDA_ARCH__ preprocessor macro to a value representing theCUDA Compute Capability. Section 6.7.4 of the NVCC documentation("CUDA Compiler Driver NVCC," included with the CUDA Toolkit)states that:

…the architecture identification macro__CUDA_ARCH__ is assigned a three-digit value string xy0 (ending in aliteral 0) during each nvcc compilation stage 1 that compilesfor compute_xy. This macro can be used in the implementation of GPUfunctions for determining the virtual architecture for which it iscurrently being compiled. The host code (the non-GPU code) must notdepend on it.

As an example,this function uses the macro to only call Compute Capability 2.0functions on devices that support them:

#if __CUDA_ARCH__>= 200
intb = __ballot();
intp = popc(b & lanemask);
#else
//do something else for earlier architectures
#endif

(Source: Stack Overflow)

CUDA CALENDAR

back to the top

October-December

- GPU Acceleration inBioinformatics

Oct. 9, 2012,London, U.K.
Sponsored by Institute for Cancer Research and NVIDIA
http://www.icr.ac.uk/research/seminar_page/index.shtml


- OpenACC Directives with PGIAccelerator Compilers (Webinar)

Oct. 11,2012
Taught by Michael Wolfe, PGI
Note: Final webinar in a 3-part series
https://www2.gotomeeting.com/register/606609858


- OpenACC Workshop (RWTHAachen University)

Oct. 11-12, 2012,Aachen, Germany
Taught by Timo Stich, NVIDIA and Sandra Wienke, RWTH AachenUniversity
http://www.rz.rwth-aachen.de/go/id/tmy?lang=en


- Nsight Eclipse: HighProductivity IDE for CUDA on Linux, MacOS(Webinar)

Oct. 12, 2012,10:00 am IST
https://www2.gotomeeting.com/register/789709786


- Portability, Scalability,Numerical Stability in Accelerated Kernels(Webinar)

Oct. 16, 2012,9:00 am PT
Speaker: John Stratton, University of Illinois atUrbana-Champaign
http://www.gputechconf.com/page/gtc-express-webinar.html


- OpenACC Workshop (NationalScience Foundation, USA)

Oct. 16-17,2012
Hosted by Pittsburgh Supercomputer Center, Natl. Inst. forComputational Sciences, Georgia Tech
Note: Telecast to 10 satellite sites around the USA
http://www.psc.edu/index.php/training/openacc-gpu-programming


- Dell HPC SolutionsWorkshop

Oct. 25, 2012,Baltimore, Maryland
Note: Includes talks by researchers, complimentary lunch
http://bit.ly/PG8apz


- GPU Accelerated Applicationsand Academic Research

Oct. 31, 2012,10:00 am PT
Speaker: Devang Sachdev, NVIDIA
https://www2.gotomeeting.com/register/206620714


- 4-Day CUDA Training Course(Acceleware)

Nov. 6-9, 2012,Houston, Texas
http://acceleware.com/training-course/16


- SC12

Nov. 10-16, 2012,Salt Lake City, Utah
http://sc12.supercomputing.org/


- GPUs in theCloud

Dec. 3-6, 2012,Taipei, Taiwan
http://gpu-cloud.cs.astate.edu/


- Many-Core DeveloperConference (UKMAC 2012)

Dec. 5, 2012,University of Bristol, UK
http://www.cs.bris.ac.uk/Research/Micro/UKMAC2012/

2013

- GPU Tech Conference

March 18-21,2013, San Jose, Calif.
http://www.gputechconf.com/

(To list an event, email: cuda_week_in_review@nvidia.com)

CUDARESOURCES

Downloads

– CUDA 5:www.nvidia.com/getcuda
– CUDA 5 survey: www.surveymonkey.com/s/CUDA5Feedback
– Nsight: http://www.nvidia.com/object/nsight.html
– CARMA (pre-register): http://bit.ly/carma_news

CUDA on the Web

– GPU Test Drive:www.nvidia.com/gputestdrive
– Learn more about CUDA: www.nvidia.com/cudazone
– Follow CUDA & GPU Computing on Twitter: www.twitter.com/gpucomputing
– Network with other developers: www.gpucomputing.net
– Stay tuned to GPGPU news and events: www.gpgpu.org
– Newsletter archive: www.nvidia.com/object/cuda-newsletter.html
– CUDA Spotlights: www.nvidia.com/object/cuda-in-action.html

ABOUT CUDA

CUDA® is aparallel computing platform and programming model invented byNVIDIA. It enables dramatic increases in computing performance byharnessing the power of the graphics processing unit (GPU). NVIDIAprovides a complete toolkitfor programming on the CUDA architecture, supporting standard computinglanguages such as C, C++ and Fortran. Send comments andsuggestions on the newsletter to cuda_week_in_review@nvidia.com

You are receivingthis email because you have previously expressed interest in NVIDIAproducts and technologies. Click here to opt in specifically to CUDA: Week inReview.

Feel free toforward this email to customers, partners and colleagues.

Please note that this message was sent to the following emailaddress: (ccole@nvidia.com)

Copyright © 2012 NVIDIA Corporation. All rights reserved. 2701 SanTomas Expressway, Santa Clara, CA 95050.

  

爱华网本文地址 » http://www.413yy.cn/a/25101014/232024.html

更多阅读

魔影工厂转码慢的问题 魔影工厂mac版

最近遇到一个挺郁闷的问题,新装的电脑,配置不低,B85-PLUS 、I5 4570、 8G内存、750ti显卡,转换视频非常慢,慢了好几倍,硬件换了个便,还是同样的问题,弄了好多天,终于发现问题出现在显卡驱动的问题,最新的驱动未必最好,用最新的驱动打不开CUDA

CUDA是什么? cuda5.0是什么

最近公司成立个专案叫什么CUDA研究,简单看了一下,公司的CUDA技术主要用于喷漆领域和图形处理.CUDA是什么CUDA是NVIDIA近来力推的东西,简单的说,CUDA可以认为是一种以C语言为基础的平台,主要是利用显卡强大的浮点运算能力来完成以往需

声明:《CUDA:WeekinReview》为网友酷不停分享!如侵犯到您的合法权益请联系我们删除