aditya deo resume

8
Society for Computer Technology and Research’s Pune Institute Of Computer Technology S.No. 27,Dhankawadi, Pune Satara Road, Pune 411043. ADITYA DEO B.E. (Information Technology) Full Name Aditya Vighnesh Deo Address E-18 Manmohan Parshwanath Society, Bibwewadi, Pune 411037. Photograph Date of Birth 17 th of August 1990 Sex Male Email [email protected] Mobile Number 9922080965 Alternate Contact Number 9604901339 PERSONAL DETAILS:

Upload: vydeo

Post on 18-Apr-2015

44 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Aditya Deo Resume

Society for Computer Technology and Research’sPune Institute Of Computer Technology

S.No. 27,Dhankawadi, Pune Satara Road, Pune 411043.

ADITYA DEO B.E. (Information Technology)

Full Name Aditya Vighnesh Deo

Address E-18 Manmohan Parshwanath Society, Bibwewadi, Pune 411037.

Photograph

Date of Birth 17th of August 1990

Sex Male

Email [email protected]

Mobile Number 9922080965

Alternate Contact Number 9604901339

PERSONAL DETAILS:

Page 2: Aditya Deo Resume

EDUCATIONAL QUALIFICATIONS:

Course SSC HSC Engineering

InstitutionN.E.S.T. Pune S.P. College Pune PICT, Pune

Year of Passing2006 2008 2012

SpecializationScience and Mathematics

Information Technology

Percentage89.73 76.33 62.59*

Class/GradeDistinction Distinction 1st Class

*Aggregate of first seven semesters.

INTERESTS:

Information Assurance and Security and Networking Theory of Computation Digital Electronics Computer Organization Operating Systems Systems Programming Artificial Intelligence Database Management Systems

COMPUTER SKILLS: Languages : C, C++(basic), SQL (basic), Visual Basic (basic), Java (basic),

CUDA (basic), UML 2.0 Software Testing: : WinRunner (Automated testing) Web designing : HTML (basic) Operating System : Windows, Linux (Ubuntu, basic)

Highly proficient programmer in C and Java

Page 3: Aditya Deo Resume

Internship at NVIDIA Corporation

Currently I am working as an intern at NVIDIA Corporation. I am a part CUDA Software Development Team. The total period of internship is 11 months, started from 11th of July 2011. We are working to enhance CUDA Visual Profiler to support intra-kernel profiling.

Background of the work being done:

CUDA (Compute Unified Device Architecture) is NVIDIA’a GPU architecture which lets one use a GPU as a general purpose co-processor. You write a heterogeneous program targeted for CUDA in the language “C for CUDA” which is an extension to standard C language with some restrictions. In C for CUDA, you write the code which you want to execute on a GPU as a separate function. The compiled code for this function is called as “kernel”.

A profiler is a tool which helps in doing posteriori analysis of a program by measuring the performance of a program run. While a program is executing a profiler collects various data about that execution. The data includes total_gpu_time, total_L1_cache_misses, total_control_flow_divergence, global_memory_throughput, static_shared_memory_per_block, etc. These collected data are analysed to find if and how the program can be optimized. Currently Nvidia’s CUDA profiling tool supports only kernel-level analysis i.e. it outputs all the parameters as aggregate for entire kernel. But this granularity is not sufficient while optimizing large and complex kernels as it hides bottlenecks inside a kernel. Hence division of the parameters at individual source and machine instruction level is desired. This is called as intra-kernel profiling.

Our work: Find mapping between CUDA C source program lines and assembly

instructions and vice-versa for optimized builds. Obtain data collected at machine instruction level by using the driver API. Compute and then display various parameters for individual CUDA C

instructions. Patch machine code to separate out interesting code sections and

compute various parameters for these sections. Display these sections and the parameters do the programmer.

Obtain dis-assembly of machine code by using dis-assembler API. Graphically show the mapping between assembly code and the source

code. Implement above mentioned functionality as a DLL and hook it up with

the driver.

Page 4: Aditya Deo Resume

PROJECT WORK:

Title About the project

Implemented IEEE paper:

Outperforming LRU with an adaptive replacement cache algorithm

Language used: CPaper Abstract: The self-tuning, low-overhead, scan-resistant adaptive replacement cache algorithm outperforms the least-recently-used algorithm by dynamically responding to changing access patterns and continually balancing between workload recency and frequency features. Caching, a fundamental metaphor in modern computing, finds wide application in storage systems, databases, Web servers, middleware, processors, file systems, disk drives, redundant array of independent disks controllers, operating systems, and other applications such as data compression and list updating. In a two-level memory hierarchy, a cache performs faster than auxiliary storage, but it is more expensive. Cost concerns thus usually limit cache size to a fraction of the auxiliary memory's size.

College Timetable Generator

Technologies used: VB 6, OracleProject Description: Was supposed to automatically generate college timetable and fill it in database database and then answer queries. But the timetable generation part got only half-completed other parts were fully functionalRole: Project Architecture Design and Back-end in SQL.

Broken Links Detector Technologies used: Java, HTMLProject Description: The purpose of the application is to help website owners maintain their sites .It begins with the link given by the user, downloads the corresponding webpage, extracts links from the webpage (if any), tests the links and updates the report. This procedure is repeated for every good link found in the user’s domain. The report is finally sent to the user. The application reduces user’s efforts to manually test all the links and also saves her bandwidth.Role: Java, HTML developer.

Page 5: Aditya Deo Resume

General Purpose Calculator

Language used: C

Project Description: The calculator can evaluate any equation in two variables x and y. The user has to input the expression through keyboard just as we write it on paper.

Sample Input: (x+2*(y-x)^7) * cos(8+(5*log(sin(6*x-2))) + x^2-y^2 + cos(tan(sin(x-y))), and values of x and y.

CSV File Analyser and Merger Utility

Language used: C

CSV is a textual file format. The Visual Profiler for Nvidia GPUs produces output in CSV format. This utility analyzes the profiler output to give

1. Summary of a run. 2. Difference between a run and the gold run.

The utility also merges results of various runs to form an aggregate CSV file.

ACHIEVEMENTS: Stood 7th in district level chess competition in 2004. Got National Merit Scholarship in 2007.

HOBBIES: Playing Tabla. Playing Badminton. Playing Chess. Reading books.