• Hi!
    I'm Amit

  • I am
    a HPC Researcher

About Me

Who Am I?

Hi I'm Amit Ruhela With over 16 years of academic and industrial research experience, I turned my passion for systematic research into a career, succeeding and advancing research in High Performance Computing. I have committed to deliver excellence and enriching the end-user experience through the best application's performance. I look forward to setting my vision and research drifts while inspiring multi-discipline teams to achieve the best outcomes on global projects.

I earned a Ph.D. degree from Indian Institute of Technology Delhi, India (top-ranked university of India in 2016) and B.Tech. degree in Computer Science from National Institute of Technology, Kurukshetra, India. I have worked with researchers from IIT Delhi and National ICT Australia (NICTA, now merged with CSIRO to form Data61) during my Ph.D. I am currently researching at Network-Based Computing (NBC) Laboratory in The Ohio State University, USA in the HPC domain. I am fortunate to work with one of the most talented, visionary, and understanding professor "Dhabaleswar K. (DK) Panda".

High Performance Computing

Computer Networks

Data Analytics

Social Computing

Cups of Coffee
Projects
Organizations
My Specialty

My Skills

Research Experiences: High-performance interconnects and protocols, Parallel Computer Architecture and algorithms, Network-based Computing, Microbenchmarks and Multi-Petascale HPC/Cloud/DL applications, Information Retrieval and Data Mining, Machine Learning

Summary: Extensive research experience spanning over 8 years in High Performance Computing (HPC), Information Retrieval, Data Mining, and Social Computing. Over 13 years of industrial research and development experience in developing world-class software. Successful in individual contributions as well as in collaborative research.

Skills: Strong problem-solving skills, Project Management Skills, An effective listener and a communicator

C/C++/Java/Python

90%

Tools(GnuPlot/TikZ/Gdb/Git/mpiP/Scalsca/Bugzilla)

100%

Message Passing Interface(MVAPICH2/Open MPI/Intel MPI/HPC-X/Spectrum MPI)

90%

HPC Applications

100%

Data Analytics(Machine Learning/Statistical Analysis/Modeling/Inference)

90%

Databases(MySQL/Oracle/DB2/ExtremeDB/NoSQL)

90%

Switching platforms: Broadcom/Fulcrum/
Winpath3

85%
Education

Education

Achievements

- Designed efficient algorithms for MPI broadcast collective communication that shows 77% performance benefits

- Improved performance of HPC applications by more than 60% with an efficient algorithm design for MPI asynchronous progress communication

- Analyzed and showcased the impact of modern interconnects (InfiniBand EDR and HDR) on the performance of HPC, Big Data and Deep Learning applications

- Developed interfaces to expose performance parameters of MPI processes to performance tools and debugging applications

- Designed and implemented profiling support to enquire and configure control variables at global and communicator levels that enable MPI implementations to tune and yield the best performance

- Developed programs to analyze the performance metrics with Xilinx based FPGA

- Coordinated software release activities of MVAPICH2 libraries

- Tuned performance of MVAPICH2 MPI libraries for multi-petascale HPC applications on several large-scaled clusters

- Extended support in writing three NSF and an XSEDE proposals

Indian Institute of Technology Delhi, India

Year : 2016

Thesis title : Analysis of the spread of information on Twitter and its application to the Internet content distribution

Advisor : Dr. Aaditeshwar Seth

Indian Institute of Science, India

Fall, 2004

Grade A

Advisor : Dr Anasuya Devi, IISc Bangalore

National Institute of Technology, Kurushetra, India

1999-2003

CGPA : 75%

Publications

Publications

Efficient asynchronous communication progress for MPI without dedicated resources. International Scholar Research Exposition, Columbus Ohio, USA, 2018

Efficient Asynchronous Communication Progress for MPI without Dedicated Resources. 6th Annual MVAPICH User Group (MUG) Meeting, Columbus Ohio, USA, Aug 06-08, 2018

Analysis of the spread of Information on Twitter and its application to Internet content distribution, 9th Inter-Research-Institute Student Seminar in Computer Science(IRISS), Goa India, Feb. 2015

Studying the characteristics of Popular users on Twitter, 8th ACM International Conference on Distributed Event Based Systems (DEBS), Mumbai India, May, 2014

The Scope for Online Social Network Aided Caching in Web CDNs, 10th i2tech OpenHouse, IIT Delhi, India, Apr. 2014

Spatio-Temporal and Events Based Analysis of Topic Popularity in Twitter, 9th i2tech OpenHouse, IIT Delhi, India, Apr. 2013

Towards the use of Online Social Networks for Efficient Internet Content Distribution (COMSNETS), Bengaluru, India, Jan. 2012

Scalable Content Distribution on the Internet for Social Networking Websites, 12th International Conference on Distributed Computing and Networking (ICDCN), Bengaluru, India, Jan. 2011

Effient asynchronous communication progress for MPI without dedicated resources, CSE 6429, The Ohio State University, USA, September 19, 2018

Analysis of the spread of information on Twitter and its application to the Internet content distribution, CSE Seminar Series, IIT Delhi, India, November 27, 2015

T3. Understanding Twitter streaming APIs and NLP tools from a developer's perspective, CSE Department, IIT Delhi, India, Feb. 2016

Experience

Work Experience

Postdoctoral Research (The Ohio State University) 2017-Present

Achievements

- Designed efficient algorithms for MPI broadcast collective communication that shows 77% performance benefits

- Improved performance of HPC applications by more than 60% with an efficient algorithm design for MPI asynchronous progress communication

- Analyzed and showcased the impact of modern interconnects (InfiniBand EDR and HDR) on the performance of HPC, Big Data and Deep Learning applications

- Developed interfaces to expose performance parameters of MPI processes to performance tools and debugging applications

- Designed and implemented profiling support to enquire and configure control variables at global and communicator levels that enable MPI implementations to tune and yield the best performance

- Developed programs to analyze the performance metrics with Xilinx based FPGA

- Coordinated software release activities of MVAPICH2 libraries

- Tuned performance of MVAPICH2 MPI libraries for multi-petascale HPC applications on several large-scaled clusters

- Extended support in writing three NSF and an XSEDE proposals

TEACHING ASSISTANTSHIP (IIT Delhi)

CSL374: Computer Networks, Instructor : Dr. Aaditeshwar Seth, Fall 2015-16

SIL801: Special Topics in Multimedia System, Instructor : Dr. Aaditeshwar Seth, Spring 2015-16

Graduate Research Assistant (IIT Delhi) 2010-2016

Achievements :

- Analyzed characteristics of OSN users and patterns of topics spread to optimize content placement strategies in web CDNs

- Proposed a novel approach of topic identification to label topics to the content of a big dataset within 15 days instead of 3000 days

- Applied techniques of machine learning (classification, clustering, curve fitting), statistical analysis (sampling, correlation, time series analysis, probability distributions), statistical modeling (data generation), and statistical inference to analyze datasets having 196M tweets, 1.47B social relations, 10M users

- Applied text-mining to pre-process user content (tokenization, stopwords, normalization and stemming), NLP to extract keywords ( entities, relations, events and topics ) from content, and Geo web APIs to parse geographical information of end-users

- Collaboration with researchers from NICTA Sydney and IIT Delhi

- Received a best presentation award in COMSNETS 2012 conference for the paper titled “Scalable content distribution for social networking websites”

Senior Researcher (C-DOT) 2003-2017

C-DOT is the premier telecom R&D center of the Government of India dedicated to developing technology products of national importance

Summary : Setting and leading the development of world class softwares that meet strategic needs of modern telecommunication, conforms to international standards ( 3GP, ITU-T and Broadband forum) and complies to stringent regulatory requirements. Well versed with complete SDLC. Timely promoted throughout career in R&D organization.

Achievements

- Developed application software for GPON project that has won tenders to provide technology backbone for $5.6 billion BharatNet project under Digital India initiative. The initiative will connect 250,000 Gram Panchayats providing them with video, voice, and data services bridging the rural-urban digital divide. G-PON project was adjudged the winner of the IESA Technovation 2015 award in most innovative product category.

- Lead development of pizza-box variant of GPON OLT to cater urban markets. Product was awarded for innovation by ELCINA in 2013-14 for its ultra-compact design.

- Comprehensively redesigned the most critical datapath module in GPON systems improving functionality, efficiency, maintainability. The design made system inter-operable resulting in development of four more variants of GPON with components sourced from different vendors.

- Designed and developed single box ONT devices that deliver all four services ie. Voice, Video, Data, and CES

- Enhanced capabilities of G-PON OLT stack by implementing L3 routing protocols which lead to a cost-effective single card solution and saved millions of CapEx of network service providers

- Designed in-memory database that improved system availability to 99.999%.

- Developed a new debugging tool to fine-grained localize datapath issues in G-PON devices that resulted in a 20% improvement in software delivery time

- Developed applications to facilitate online debugging of system installed in remote areas saving personnel man-hours and travel costs.

Visiting Researcher (Data61) Sept-Oct 2012 and July-Aug 2013

Achievements

- Proposed an efficient algorithm to extract the event’s dynamics by analyzing 6M popular, medium popular, and niche topics of Twitter OSN

- Design and implemented a three-phase discrete event simulator to benchmark and compare the performance of different caching algorithms against each other

- Addressed the tractability of tracking millions of user-generated contents to expose trend in the spread of topics

Services

Services

Referee

JPDC 2019 : Journal of Parallel and Distributed Computing

ISC 2019 : ISC High Performance - Programming Models & Systems Software

EuroMPI 2019, 2018 : 26th European MPI Users' Group Meeting

IPDPS 2018 : International Parallel and Distributed Processing Symposium

OpenSHMEM 2018: Fifth Workshop on OpenSHMEM and Related Technologies.

TACO 2018 : ACM Transactions on Architecture and Code Optimization

Judge

Ohio Academy of State Science Day
May 2019

Volunteer

Swachh Bharat Mission:
2017