Ezhilmathi Krishnasamy

Ezhilmathi Krishnasamy

I am a postdoctoral researcher at Luxembourg University, working with the Parallel Computing and Optimization group. My research interests are scientific and quantum computing.

Activity

  • Hi, this was not supposed to be there, but it has now been removed.

  • Hi, that file gives an idea about how to create different dimensions of a thread. And how to convert them into suitable forms, for example, 1D or 2D. The performance of the thread block varies depending on the type of computation/problem. As a user, in the end, you have to choose a suitable thread block for your application. There is no constant thread block…

  • Hi, thanks for the finding error, now it has been changed.

  • Hi! thanks for your feedback, we will add more visuals in the next run!

  • Hi! thanks for your feedback, we will add more visuals in the next run!

  • Hi! thanks for your feedback, we will add more visuals in the next run!

  • Hi! thanks for your feedback, we will add more visuals/info in the next run!

  • Hi! thanks for your feedback, we will add more visuals in the next run!

  • Hi! thanks for your feedback!

  • Hi, as we have discussed earlier, CUDA supports up to 3D dimension threads. However, here we only need a 1D thread to do the vector operations. Creating 2D thread blocks in CUDA is somehow efficient for computation. That is why we have shown here 1D and 2D thread creation for 1D operation (vector operation).

  • Hi, CUDA threads can be created in the form of 1D, 2D, and 3D. Depending on your application, you might want to use it. However, you also need to consider the CUDA occupancy for the better performance of the threads. Thanks for your feedback, we will add more examples in this section.

  • Hi, Thanks! it has been fixed now. In Fig3, the sub-index m refers to, the link continues. For example, 1,2,3,…m memory for 1,2,3,…p processor.

  • Hi, Thanks! it has been fixed now.

  • Hi! we will try to improve the next run! and we will process all the information that we get from this run.

  • We are presently dealing with green energy (renewable energy). Basically using the super computer for scientific applications is always beneficial. For example, using wind tunnel testing for car modeling or airplane modeling required a lot of electricity/manpower/time consumption.