CUDA Online Resources

[Saiyedul Islam]

Hi all,

Here are few very useful online resources for learning CUDA.

Resource 1 - CUDA Book in BITS Library

Recently, I found that our library's online portal has the first ever book published on CUDA.

"Programming Massively Parallel Processors - A hands-on approach" by David Kirk and Wen-mei Hue.

David Kirk is a scientist in Nvidia and Hue is professor at UIUC (as far as I know)

They both are pioneer in the field of GPU programming.

To access the book, go to https://ebooks.bits-pilani.ac.in

After logging in using your account, serach for "cuda". This will be the 3rd result.

Resource 2 - Online CUDA Course

Dr. Hue is offering  a course on Heterogeneous Parallel Programming

This will be a 6 weeks long course which will start in October and will cover CUDA/OpenCL and MPI.

After successful completion of course you will get a certificate signed by him.

Resource 3 - CUDA Webinar by Nvidia

For many years, Nvidia used to offer online webinars of CUDA but they used to be in US time.

From last year, they have started offering webinars in India time also.

Next such webinar is on this Friday (21st Sep, 2012).

Timings - 10:00 am to 11:00 am (IST)

Register yourself at - Introduction to GPU Computing & CUDA Architecture

Though it will be a small talk but you will get to hear what Nvidia wants to say about their baby (CUDA)

Also, if large number of people attend this webinar by mentioning their organization name as BITS Pilani, they will definitely notice our name.

I hope that you will use these resources to the fullest.

PS - Due to Schemcon on Friday (21st Sep, 2012), we are cancelling CUDA 4th session that was planned on the same day.

       New date and timings will be posted very soon.

Regards,

Saiyedul Islam

[Saiyedul Islam]

Hi guys,

This is a great article for .NET lovers who can use GPUs to accelerate their codes.

The website GPUScience recently wrote about CUDAfy, a set of libraries and tools for programming GPUs within the Microsoft .NET framework. The post refers to an interesting article by John Michael Hauck on how to transfer CPU code to the GPU using the Traveling Salesman problem as an example. 

 

http://gpuscience.com/code-examples/cudafy-me-traveling-salesman-problem-with-cuda-from-c/

Cheers,

Saiyedul

[rohini jayanthi]

Any lecture on Linux after mid-sem?