Evaluating the Performance of Interpreters Over Various Branch Predictors

Interpreters introduced a high number of indirect branches. In this project we analyze the performance of three popular interpreters (Python, Java, JavaScript) over various branch predictors. All the hardware counters data collected using OProfile, an open source project that includes a statistical profiler for Linux systems, capable of profiling all running code at low overhead. (http://oprofile.sourceforge.net/news/)

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes. See deployment for notes on how to deploy the project on a live system.

Prerequisites

You need to install:

OProfile (Ubuntu 16.10)

    $ sudo apt-get install oprofile

OProfile (Centos 7)

    $ sudo yum install oprofile

Install Python3.6 (Ubuntu 16.10)

    $ sudo apt-get update
    $ sudo apt-get install python3.6

Install Python3.6 (Centos 7)

    $ sudo yum -y install python36u

Install python-pip

	$ wget https://bootstrap.pypa.io/get-pip.py
	$ sudo python3.6 get-pip.py

Operf

    $ sudo python3.6 -m pip install perf	

Install Rhino (Ubuntu 16.10)

    $ sudo apt-get install rhino

Install Rhino (Centos 7)

    $ yum install rhino

Running the tests

All the tests must run inside bechmarks directory

    $ cd benchmarks/

Run the benchmarks

    $ ./1.runNTimes.sh

Parse the results

    $ ./2.parseResutls.sh

Plot the results

    $ ./3.runGNUGraphs.sh

Interpreted Languages :

1. JavaScript : https://www.javascript.com/
2. Java       : https://www.java.com/en/
3. Python     : https://ww.python.org/	

Hardware Architectures

1. Intel Core 2 Duo
2. Intel Nehalem(2008)
3. Intel Ivy Bridge(2012)
4. Intel Haswell(2013)
5. AMD64 family15h

Benchmarks

1. JavaScript    : https://github.com/chromium/octane.git
2. Java          : https://github.com/dacapobench/dacapobench.git
3. Python        : https://github.com/python/performance.git