In order to popularize high performance computing (HPC) among the student community in India, the NSM Nodal Center for Training in HPC and AI at IIT Goa invites applications for internships as listed in the table below. These will be Remote Internships. There is an internship stipend of Rs. 8,000 per month, pro-rata basis, for a maximum of three months. However, all selected interns will be paid their internship stipends only upon successful completion of the internship, as certified by the internship guides. These internships are supported by the National Supercomputing Mission (NSM) of the Govt. of India together with CDAC.

Internship Period

Minimum 2 months and maximum 3 months.
Should begin by 1st June 2022 and end by 31st August 2022. Selected intern and guide are free to choose any period in this range based on mutual convenience.

Eligibility (including COVID-19 impact on education)

Self motivated and passionate engineering and science background students can apply as mentioned in the table of projects shown below. However, considering the impact of COVID-19 on education, students graduated in 2021 and 2022 and waiting to join a university abroad for higher studies (MS, PhD etc.) are also eligible to apply. They will need to upload the offer of admission and other supporting documents as a single file as mentioned in the application form. Note that 2nd year and above means completed at least two years in a UG program. 2nd year onward means completed the first year in a UG program.

How to apply

Please fill up the online application using this link
Application Period: 4th May 2022 to 11th May 2022
There is no application fee. Please write to concerned guides in case of any project related queries.
General queries can be sent to Sunit Fulari.

Selection Procedure

Based on CPI/CGPA/Marks, prior projects, technical skills etc. Some guides may conduct interviews. All applicants will be informed about the outcome of their applications.
All applicants and interns have to abide by the following Terms and Conditions. Submission of your application means that you abide by them:

  1. Internships are available only to students as mentioned under different projects. All internships are remote internships.
  2. Applicants should upload:
    1. A scanned copy of their college/university/institute ID card clearly showing name, degree program, validity period, photograph, address etc.
    2. A scanned headshot of theirs
    3. PDF copy of offer of admission abroad and any other supporting document, if the applicant graduated in 2021 or is going to graduate in 2022
    4. Their latest mark sheets showing CPI/CGPA/Percentage
    5. Curriculum Vitae (CV) with a list of relevant courses, technical skills, projects done, github link (if any) etc.
  3. Interns may be provided remote access to servers etc. at IIT Goa by project guides, if needed. All interns are also expected to have a laptop/desktop (at least i7, 8 MB RAM, 512 GB HDD) or access to such a machine for their work.
  4. All interns shall abide by all rules governing usage of IIT Goa servers etc. and shall not use it for any activity other than those related to the internship. Internships will be terminated in case of any violation.
  5. Internship stipend shall be paid only at the end of internship upon certification by the project guides.
  6. The decision of NSM Nodal Center on all matters is final.

Project Number Name of the guide School Brief description of the project (max 200 words) Target students (UG, PG, Year of study etc.) Skills expected in an intern selected for the project Files, if any
NSMGOA1 Prof. B K Mishra School of Mechanical Sciences This project will study sphere packing using Monte Carlo method UG (Second year and above) and PG (MS/MTech/MSc) Basic understanding of sphere packing, parallel programming
NSMGOA2 Prof. Y Sudhakar School of Mechanical Sciences This project focuses on developing basic Computational fluid dynamic (CFD) parallel programs on GPUs. The aim is to write codes with detailed comments and necessary documentation. These will be published in open source repositories to serve as a starting point for students/researchers to learn GPU programming in the context of CFD. Additional details can be found in the attachment. UG (Second year and above) and PG (MS/MTech/MSc) Knowledge of numerical methods, discretisation of the Poisson Equation using finite difference methods, C/C++ programming pdf
NSMGOA3 Prof. Neha Karanjkar School of Mathematics and Computer Science Sitar (https://nehakaranjkar.github.io/sitar/) is an open source simulation framework we have developed, designed for performing fast scalable simulations on many-core systems. The advertised project is targeted to design and perform a set of experiments to evaluate the performance and scalability of Sitar on a many core system and compare the performance with alternate frameworks such as SimPy and SystemC. 2nd year onwards UG or PG (MS/MTech/MSc) Good familiarity with C++, Python and Linux (shell or Python scripting), experience in projects involving object-oriented programming, basic familiarity with OpenMP. Knowledge of discrete-event simulation algorithms is desirable. pdf
NSMGOA4 Prof. Ashish Bhateja School of Mechanical Sciences A comprehensive understanding of the flow of pedestrians through a door is essential for designing pedestrian facilities, ensuring an uninterrupted and safe egress of pedestrians in highly competitive scenarios. An understanding of such flows becomes important while analysing a variety of social scenarios involving highly crowded environments. This work intends to examine pedestrian motion through a door in various situations by employing computations based on a discrete element technique. Initially, the plan is to develop a serial code, which will be made parallel using MPI (message passing interface) for modelling large systems comprising thousands of people. UG (Second year and above); PG (MS/MTech/MSc) Good programming skills. Working knowledge of FORTRAN. pdf
NSMGOA5 Prof. Sharad Sinha School of Mathematics and Computer Science Open source tutorial on Heat Sink Design for Semiconductor Chip Packages (like BGA) using OpenFOAM, Salome and Paraview. Novel architectures other than the fin-based heat sink architectures will need to be explored in the tutorial once the fin-based heat sink architecture tutorial is complete. Preparation of a survey on various existing heat sink designs for chip packages. UG (Second year and above); PG (MS/MTech/MSc) At least basic understanding of OpenFOAM. The rest can be learned during the internship. Basic understanding of heat sink, heat flow, chip packaging etc. Open to students from CSE/ME/EE
NSMGOA6 Prof. Sharad Sinha School of Mathematics and Computer Science Optimizing for speed and parallelizing the implementation of bioinformatics software AUGUSTUS. After sequencing the genome (DNA) of any species, molecular biologists are tasked with figuring out which sections of the genome encode genes (exons) and which sections are non-coding (introns). This task is called gene prediction. AUGUSTUS is a gene-prediction software (part of the BRAKER pipeline) that can locate genes by looking at just the genome sequence (ab initio). If available, it can also incorporate “extrinsic hints” from protein / RNA sequencing of that species. AUGUSTUS usually takes a long time for gene prediction, even when running on High-Performance Computers (HPCs). Two reasons for this are poor parallelization and performance bottlenecks. Hence, the goal is to improve the implementation of AUGUSTUS to make it faster and more suitable for HPCs. UG (Second year and above); PG (MS/MTech/MSc) Program parallelization, multi-threading, concurrency, code optimization techniques, process synchronization, OS concepts, basics of molecular biology, learning on the job! pdf
NSMGOA7 Prof. Lok Pati Tripathi School of Mathematics and Computer Science The project is on the comparison of different pricing methods for multi-asset options in terms of their accuracy and efficiency. Main focus will be on (1) the probabilistic approach: Monte Carlo method, and (2) the deterministic approach: approximating the solution of high dimensional Black-Scholes partial differential equation by using grid based methods and Physics Informed Neural Networks (PINNs). UG (3rd year and above); PG (MS/MTech/MSc) Basic probability theory, partial differential equations, and familiarity with the Python programming language