www.gpgpu.org

Employment Postings :: Point Cloud Library Opportunity (Google Summer of Code)

Author: dom
Subject: Point Cloud Library Opportunity (Google Summer of Code)
Posted: Fri Mar 23, 2012 8:45 am (GMT 0)
Topic Replies: 0

[posted on behalf of Alexandru Eugen Ichim]

We are very proud to announce that the Point Cloud Library has been accepted for the second time in the Google Summer of Code program. This is a great opportunity to have new people joining the thriving community of open-source 3D perception.

The webpage where you will find our list of ideas is http://www.pointclouds.org/gsoc2012/ideas.html. We are very flexible with the project ideas, and we are accepting interesting and solid (!!!) project proposals on our official event mailing list:gsoc2012@pointclouds.org or for more direct communication on IRC: irc.oftc.net, channel #pcl.

In order to apply, please follow the instruction on the official GSoC website (http://www.google-melange.com/gsoc/org/google/gsoc2012/pcl)and our custom student application requirements at http://www.pointclouds.org/gsoc2012/application.html.

A few words about PCL:

The Point Cloud Library (PCL) is a standalone, large scale, open project for point cloud processing. A point cloud is a data structure used to represent a collection of multi-dimensional points and is commonly used to represent three-dimensional data such as the output of a stereo camera, 3D scanner, or time-of-flight camera. With the advent of new, low-cost hardware such as the Microsoft Kinect or Asus XTionPro 3D cameras and continued efforts in advanced open source 3D point cloud processing, 3D perception gains more and more importance in many fields.

The PCL library contains numerous state-of-the art algorithms including filtering, feature estimation, surface reconstruction, registration, model fitting, segmentation, tracking, recognition, and many more. These algorithms can be used, for example, to filter outliers from noisy data, stitch 3D point clouds together, segment relevant parts of a scene, extract keypoints and compute descriptors to recognize objects in the world, and create surfaces from point clouds and visualize them -- to name a few.

The Point Cloud Library aims to unite the field of point cloud processing. By providing an extensible framework for all the geometric algorithms necessary for 3D perception, PCL enables developers to create applications limited by their imaginations, rather than their 3D geometric knowledge.

Employment Postings :: Intuitive Surgical Research Intern

Author: dom
Subject: Intuitive Surgical Research Intern
Posted: Sat Mar 03, 2012 9:02 am (GMT 0)
Topic Replies: 0

[posted on behalf of Prash Chopra]

Intuitive Surgical has an immediate opening for internship in the area of medical imaging. We are seeking a talented individual to participate in cutting edge research and development using medical imaging.

Expected contributions include

- Development and unit testing of medical imaging and geometry processing algorithms.
- Development of GUI for integration testing of imaging algorithms.
- Software/hardware acceleration of image analysis pipelines.

Requirements:
- Passion for developing technologies aimed at improving the lives of millions of people.
- MS/PhD students or professionals in Computer Science/ Electrical Engineering/Computer Graphics areas preferred. Other exceptional candidates will also be considered. Prior industrial experience highly desired.
- Desired knowledge set: Solid working knowledge of GPU based medical image analysis algorithms (e.g. segmentation, registration, geometry/mesh processing), GUI development. Hands-on experience in one or more of these areas is strongly preferred.
- Desired skill set: C/C++ on Linux and Windows, GPU based image/geometry analysis, itk/vtk, OpenGL, pixel/vertex shaders, CUDA, GLSL, Cg, GUI development in Qt.
- Ability to perform fast prototyping and testing iterations and thinking outside the box to solve practical problems.

Start time: immediately; both full-time and part-time are considered.

Period: typically 3 months of full time; extension will be considered.
Please apply with a cover letter and resume here:

Online: http://intuitivesurgical.hodesiq.com/job_detail.asp?JobID=2900801

Direct: Prash Chopra
Intuitive Surgical Inc.
1266 Kifer Road Sunnyvale, CA 94086
Prash.Chopra@intusurg.com

Intuitive Surgical (www.intuitivesurgical.com), located in Sunnyvale, CA, is engaged in designing, manufacturing and marketing robotic surgical systems that are revolutionizing surgery. Its da Vinci Surgical system includes six manipulator arms with a total of 41 degrees of freedom, along with a stereo endoscope and 3D video display. With over 2000 installations worldwide, surgeons perform over two hundred thousand minimally-invasive surgeries per year. The da Vinci system represents an outstanding platform for the development and application of new technologies for surgery.

General GPGPU Discussion :: The GPGPU.org forums are locked

Author: admin
Subject: The GPGPU.org forums are locked
Posted: Wed Feb 29, 2012 1:09 am (GMT 0)
Topic Replies: 0

We have locked down the GPGPU.org forums due to excessive spam that we have been unable to control, and due to reduced traffic in the forums.

If you have received a spam message from this forum recently, please accept our apologies -- we have disabled private messaging so this shouldn't happen again.

If you are looking for answers about GPGPU, CUDA, or OpenCL, we recommend checking those tags on Stack Overflow: http://www.stackoverflow.com

For NVIDIA-specific questions, check the NVIDIA forums:
http://forums.nvidia.com/index.php?showforum=62
or
http://forums.developer.nvidia.com/

Employment Postings :: Two positions: Bioinformatics, Univ. Mainz, Germany

Author: dom
Subject: Two positions: Bioinformatics, Univ. Mainz, Germany
Posted: Sun Nov 27, 2011 10:55 am (GMT 0)
Topic Replies: 0

Two positions as Postdoc or PhD Student as a Research Associate

Location: Mainz, Germany
Application Deadline: 31-Dec-2011

The Chair for Parallel and Distributed Architectures at the Johannes-Gutenberg University (JGU) Mainz conducts research in the areas of High Performance Computing (HPC) and its application to Bioinformatics: http://www.staff.uni-mainz.de/schmi033/. Examples of active research areas are:

Utilization of accelerator architectures (e.g. GPUs and CUDA, FPGAs)
Design of scalable algorithms and data structures for bioinformatics
Hybrid HPC architectures (e.g. GPU clusters)
Design of parallel and distributed algorithms for „Next Generation Sequencing"

Besides conducting research in one of the listed research areas, active participation in teaching activities (e.g. conducting tutorials) is also required. Requirements Applicants for these positions must have a Master’s degree (for the Research Associate position) or a PhD degree (for the Postdoc position) in computer science, bioinformatics, mathematics, or a related subject. The applicant should have an excellent background in algorithm design and excellent programming skills (preferably in C/C++). The salary scale is TVL-13 (Wissenschaftlicher Mitarbeiter). Please send your application with the usual documents (CV, list of publications, softcopies of certificates, Names of two references) by email to Professor Bertil Schmidt (bertil.schmidt@uni-mainz.de).

Employment Postings :: GPU Software Developer position, Univ. of Strathclyde, UK

Author: dom
Subject: GPU Software Developer position, Univ. of Strathclyde, UK
Posted: Thu Nov 10, 2011 3:27 pm (GMT 0)
Topic Replies: 0

(Moderator's note: Posted on behalf of user spelai)

UNIVERSITY OF STRATHCLYDE
DEPARTMENT OF MECHANICAL & AEROSPACE ENGINEERING
GPU Software Developer
(Fixed term - initially for a period of 6 months)
£29,099 - £33,734 per annum

A consortium of three UK institutions — the University of Strathclyde, the University of Warwick, and the Science and Technology Facilities Council — is seeking to appoint a Software Developer to implement molecular dynamics simulation techniques on a GPU platform.

To be considered for this post, you must have a strong computer, engineering, mathematical or physical science background gained through a good honours degree/MSc (or equivalent qualification) and/or considerable work experience in a related subject area.

Additionally, you will have experience in one or more of the following areas:
- applying GPUs to scientific/engineering problems;
- molecular dynamics on GPUs;
- computational fluid dynamics on GPUs.

For more information please visit http://vacancies.strath.ac.uk, job reference number JA/168/2011.

Closing date: 1st December 2011

Programming Help :: arithmetic precision in fragment shaders

Author: BrianSharpe
Subject: arithmetic precision in fragment shaders
Posted: Wed Oct 12, 2011 12:59 am (GMT 0)
Topic Replies: 0

Hi guys

So I'm working on some stuff atm which requires random numbers within a pixel shader.
I'm aiming for DX9 hardware so integer hashing is not an option.
I've opted for a floating point hash from the x,y integer pixel coordinate of the form

hash = mod( x*x*y*y, SOMEPRIME ) / SOMEPRIME

I seem to be getting good results from it. ( I can clamp and offset the domain to avoid the reflection along the x=y diagonal line )

Thing is, the x*x*y*y operation pushes the floating point integer coordinate way past ( 2 << 23 ). So the floating point number starts to loose precision.
eg, x = 99.0 y = 99.0 99.0^4 > 2.0^23

So my question.
Can I expect this function to produce the same results across different cards? ( I think I read in the GPGPU faq that ati perform arithmetic in 24 bits!? rather than nvidia that does it in 32? )

thoughts?

ps: I've also posted in gamedev.net. But didn't receive the answers I was after. What I'm after is some kind of take on what to expect from fp precision on modern cards. ( most internet resources are like... pre-2005 )

Thanks all!
Ren

Advanced Development :: RE: Ambient Occlusion on GPUs in image space

Author: townie
Posted: Mon Oct 10, 2011 12:54 pm (GMT 0)
Topic Replies: 31

Also, the OBJ file doesn't have proper normal info, and that's why there's aliasing issues with the screenshots. (I just use the cross product of two edges as the triangle's normal).

Programming Help :: Issue Rendering to GL_RGBA32F_ARB FBO

Author: rmonette
Subject: Issue Rendering to GL_RGBA32F_ARB FBO
Posted: Wed Oct 05, 2011 1:45 am (GMT 0)
Topic Replies: 0

Having some issues rendering to a floating point FBO.

It seems like my data is being corrupted or misinterpreted somehow and I get results such as: http://i.imgur.com/1WjpF.png

The correct rendering would be grey for the object.

I'm assuming its OK to have normal blending modes enabled on a floating point FBO?

Doing the usual setup but I cannot find this bug!

Code:

GLuint createTexture(const int width, const int height) {

GLuint handle;

GL( glGenTextures( 1, &handle ) );
GL( glBindTexture( GL_TEXTURE_2D, handle ) );
glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_NEAREST);
glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_NEAREST);
glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, GL_CLAMP_TO_EDGE);
glTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, GL_CLAMP_TO_EDGE);
glTexParameteri(GL_TEXTURE_2D, GL_GENERATE_MIPMAP, GL_TRUE);
GL( glTexImage2D( GL_TEXTURE_2D, 0, GL_RGBA32F_ARB, width, height, 0, GL_RGBA, GL_FLOAT, 0 ) );

return handle;

}

void GPUParticleRenderer::createSingleSampleBuffer() {

singleSampleTarget.textureId = createTexture( _frameSpec.getImageSize().x, _frameSpec.getImageSize().y );

GL( glGenRenderbuffersEXT( 1, &singleSampleTarget.depthBufferId ) );
GL( glBindRenderbufferEXT( GL_RENDERBUFFER_EXT, singleSampleTarget.depthBufferId ) );
GL( glRenderbufferStorageEXT( GL_RENDERBUFFER_EXT, GL_DEPTH_COMPONENT24, _frameSpec.getImageSize().x, _frameSpec.getImageSize().y ) );

GL( glGenFramebuffersEXT( 1, &singleSampleTarget.fboId ) );
GL( glBindFramebufferEXT( GL_FRAMEBUFFER_EXT, singleSampleTarget.fboId ) );
GL( glFramebufferTexture2DEXT( GL_FRAMEBUFFER_EXT, GL_COLOR_ATTACHMENT0_EXT, GL_TEXTURE_2D, singleSampleTarget.textureId, 0 ) );
GL( glFramebufferRenderbufferEXT( GL_FRAMEBUFFER_EXT, GL_DEPTH_ATTACHMENT_EXT, GL_RENDERBUFFER_EXT, singleSampleTarget.depthBufferId ) );

if(glCheckFramebufferStatusEXT(GL_FRAMEBUFFER_EXT) != GL_FRAMEBUFFER_COMPLETE_EXT) {
EC_LOG_ERROR("buffer not complete");
}
else {
EC_LOG_INFO("buffer ok");
}

}

Reading back to the CPU in the normal manner:

Code:

glBindTexture(GL_TEXTURE_2D, singleSampleTarget.textureId);

glGetTexImage(GL_TEXTURE_2D, 0, GL_RGBA, GL_FLOAT, data.get());

Any ideas would be very appreciated!

General GPGPU Discussion :: RE: simplifying radicals calculator needed

Author: tomhanks
Posted: Fri Sep 30, 2011 7:59 am (GMT 0)
Topic Replies: 1

Ya I will surely help you.
@ the link given below this is a tool for solving your radical expressions quickly and accurately

http://calculator.tutorvista.com/math/449/simplifying-radicals-calculator.html

General GPGPU Discussion :: simplifying radicals calculator needed

Author: kanishkporwal
Subject: simplifying radicals calculator needed
Posted: Fri Sep 30, 2011 7:55 am (GMT 0)
Topic Replies: 1

I need a simplifying radicals calculator to help me completing my homework as I really suck at math could anyone please help me out.

Employment Postings :: PhD/MS positions at UW-Milwaukee

Author: rmjdsouza
Subject: PhD/MS positions at UW-Milwaukee
Posted: Tue Sep 27, 2011 10:33 pm (GMT 0)
Topic Replies: 0

There are several open positions (PhD/Masters) available in the CSSL starting immediately.

The following background would be desirable

Background in numerical methods (PDEs), particle-based methods (particle-in-cell, agent-based methods), finite elements
Solid programming skills. Parallel programming skills is a plus. GPUs, multi-core CPUs, cluster computing, MPI, Open-MP, CUDA, OpenCL
Computer graphics

Duties include researching new algorithms, literature review, code development, writing scholarly publications, and aiding the PI in writing grant proposals. Email your CV to dsouza@uwm.edu.

Visit our website at http://cssl-uwm.org

Programming Help :: Cuda linear least squares GPL

Author: aero
Subject: Cuda linear least squares GPL
Posted: Tue Sep 27, 2011 3:39 pm (GMT 0)
Topic Replies: 0

I would like to solve linear least squares using GPU but I dont know where to start. The only condition is Im supposed to use open source code.
any idea where I can get that?

Getting Started with Brook :: RE: Shared library on Windows

Author: Obershnaus
Posted: Mon Sep 26, 2011 11:53 am (GMT 0)
Topic Replies: 5

cassey37 wrote:

I have the same problem and still don't understand what is going on. Someone could explain me what exactly I am to do, if you please

if you have the same problem, then try to modify the common.mk file as described above. at least it worked for the original problem. if no - think about starting new thread with full description of your problem.
_________________
mobile application development companies

Sh Discussion :: RE: Genetic Programming with SH

Author: Wyatt
Posted: Mon Sep 26, 2011 9:16 am (GMT 0)
Topic Replies: 1

I have... And what exactly are you interested in?
_________________
google android development

General GPGPU Discussion :: RE: How to solve a linear system?

Author: kanishkporwal
Posted: Mon Sep 26, 2011 8:13 am (GMT 0)
Topic Replies: 7

Solving linear equations is not that hard problem
the problem is to convert the language into equations.

You can try this link for the information you are looking for.

http://math.tutorvista.com/algebra/linear-equations.html