Mounika Ponugoti

Mounika Ponugoti

PhD Candidate in Computer Architecture

Contact Me

Research Interests

Computer Architecture, Embedded Systems, Hardware Security, GPGPU Computing, Energy Efficiency, and Performance Profiling of Computer Systems

Publications

Enabling On-the-Fly Hardware Tracing of Data Reads in Multicores

  • Published in ACM Transactions on Embedded Computer Systems, Vol. 18, No. 4, May 2019
  • Introduces the hardware compression technique which exploits cache coherence protocol states and first access tracking bits

Read Now

Battle of Compilers: An Experimental Evaluation Using SPEC CPU 2017

  • Published in IEEE SoutheastCon, April 11-14, 2019 Huntsville, AL, USA
  • The main focus of this paper is to experimentally evaluate Intel Parallel Studio, GNU, and LLVM compilers from the user perspective using SPEC CPU2017 benchmark suite

Read Now

Master’s Thesis: Techniques for Capturing and Filtering Data Value Traces in Multicores

  • This basically includes the in-depth details of ICCD and CASES papers

Read Now

Exploiting Cache Coherence for Effective On-the-Fly Data Tracing in Multicores

  • Published in IEEE International Conference on Computer Design (ICCD’16), October 3-5, Phoenix, AZ, USA
  • Introduces the hardware trace compression technique which utilizes existing cache coherence protocol states and single bit per cache block to get the better compression ratio when data is shared in multicores

Read Now

On-the-Fly Load Data Value Tracing in Multicores

  • Published in IEEE International Conference on Compilers, Architectures and Synthesis of Embedded Systems (CASES’16), October 2-5, Pittsburgh, PA, USA
  • Introduces the hardware trace compression technique which utilizes first access tracking bits in multicores

Read Now

Work Experience

Graduate Teaching Assistant - The University of Alabama in Huntsville (Jan 2017 - Present)

Embedded Systems Laboratory

  • Teaching MSP340 assembly programming, C/C++ programming, and I/O interfacing
  • Designing new assignments, grading students work, and mentoring projects.

VLSI Design Laboratory

  • Teaching ASIC design using Verilog and Synopsys EDA tools (vcs, dc, icc, and others)
  • Grading and mentoring students projects

Programming Languages Laboratory

  • Teaching C++ programming, grading, and mentoring students projects

Digital Logic Design Laboratory

  • Teaching Verilog and Quartus-II to write and verify simple designs on Intel DE10-Lite
  • Grading and mentoring students projects

Graduate Research Assistant - The University of Alabama in Huntsville (Aug 2015 - Dec 2016)

  • Developed C++ simulator to filter hardware traces in multicores using cache coherence protocol and/or first access tracking mechanism
  • Performed architectural simulations using Multi2Sim simulator
  • Developed dynamic profiling tools using Intel’s Pin (some are here)
  • Measured performance and power profiling using Perf, Likwid, and Wattsup tools
  • Developed hardware accelerators in Verilog on DE2-115 SoC - NIOS II

Intern - RCI (Research Centre Imarat, India)(Jan 2014 - Apr 2014)

  • Worked in a project to generate phase-coded waveforms using direct digital synthesizer to increase the range and resolution of RADAR

Course Projects

Embedded platforms are ready for supercomputing?

  • The main goal of this project is to evaluate whether embedded platforms are ready for supercomputing or not
  • For this, I have measured energy efficiency, performance, and cost per performance of Core-i7 and Jetson-TK1 using Perf, Likwid, and Wattsup tools while running Parsec-3.0 and NAS parallel benchmarks

Floating point adder and subtractor using Verilog

  • ASIC module for floating point addition and subtraction is designed using Verilog and verified using Synopsys tools
  • This module can add and subtract two floating point numbers when the exponent is same

Download Now

Solving inverse boundary heat equation on GPGPU using CUDA

  • The main goal of this project is to solve the inverse boundary heat equation to determine the diffusivity matrix of the sheet using particle swarm optimization and CUDA on GPU.

Voice authentication module for Linux

  • Sudo utility is modified to use user's voice for authentication
  • While training, features of the user voice are collected and stored in a file to utilize in verification process

Download Now

Evaluation of Machine Learning Algorithms in Network Intrusion Detection

  • In literature, network intrusion detection using machine learning algorithms are evaluated using KDD CUP99 data set
  • However, this dataset has some problems. So, my goal is to verify the effectiveness of these algorithms using NSL-KDD (filtered KDD CUP99) data set

Performance Evaluation of Hypervisors

  • The main goal of this project is to evaluate the performance of bare-metal and hosted hypervisors, VirtualBox, VMware, and Hyper-V on host machine running windows and Linux OS with windows and Linux as guest OS.
  • The performance is evaluated with SPEC CPU2006 and IOzone benchmarks.