COTSON: INFRASTRUCTURE FOR FULL SYSTEM SIMULATION



Get COTSon

SVN Instructions
SF Project Page
Browse Sources
Download Page
Installation Guide
License Terms

Support

Mailing List
COTSon Wiki
Bugs & Features

Welcome to the COTSon Full-System Simulation Platform

COTSon is a full system simulation infrastructure developed by HP Labs to model complete computing systems ranging from multicore nodes up to scale-out clusters with complete network simulation. It is a pluggable architecture, in which end users can can develop their own modules, thus providing a very flexible simulation platform for a wide variety of modeling needs. The functional emulation is based on AMD's SimNow and supports x86 and x86_64 platforms.

There are tons of simulators out there, so why a new one? COTSon is not just another simulator, it is a simulation infrastructure. Our holistic approach targets the whole system at once, because we believe that understanding today's and tomorrow multicore and multithreaded architectures requires taking the whole system into account, including devices and the operating system.

As a design principle, COTSon trades off accuracy for speed and viceversa, dynamically allowing end users to determine the interesting parts of their application, and enables large space explorations at high simulation speeds. The primary expected use of COTSon is as a modeling tool to collect performance (or energy) metrics for decision support.

By publicly releasing the tool in Jan 2010 under the MIT Open Source License, we hope COTSon can become the simulation platform of choice for modeling and simulation of next-generation computing systems. If you use COTSon for your research, please use the following reference

Argollo, E., Falcón, A., Faraboschi, P., Monchiero, M., and Ortega, D. 2009.
COTSon: infrastructure for full system simulation.
SIGOPS Oper. Syst. Rev. 43, 1 (Jan. 2009), 52-61

@article{cotson-osr09,
author = {Argollo, Eduardo and Falc\'{o}n, Ayose and Faraboschi, Paolo and Monchiero, Matteo and Ortega, Daniel},
title = {COTSon: infrastructure for full system simulation},
journal = {SIGOPS Oper. Syst. Rev.},
volume = {43},
number = {1},
year = {2009},
issn = {0163-5980},
pages = {52--61},
doi = {http://doi.acm.org/10.1145/1496909.1496921},
publisher = {ACM},
address = {New York, NY, USA},
}

COTSon Repository
The recommended way to get cotson is by checking out the 'trunk' of the development tree. It contains the program sources, and an extensive set of examples. Please note that you need to separately get AMD's SimNow. We  provide periodic snapshots, but unless you have a specific reason to continue an older version, the best is to always get the latest.
For svn access:

svn co https://svn.code.sf.net/p/cotson/code/trunk cotson

If you want to help the COTSon project and contribute your changes to the main repository, please send a request to join this project


COTSon Tutorial at MICRO-41
A joint AMD/HP tutorial was held at MICRO-41 (Lake Como, Italy, Nov 2008).
Part I describes the internals of AMD’s SimNow™, the functional simulator of COTSon. AMD's SimNow software is a fast and configurable x86 and AMD64 dynamically-translating platform simulator. With SimNow, users can connect complex software models to form a PC platform emulation environment. SimNow emulates AMD Athlon™ 64 and AMD Opteron™ multiprocessor systems that run commercial operation systems and applications, including BIOS and device driver development, OS testing, and evaluation of new ISA extensions
Part II describes the way in which COTSon is connected to SimNow. A challenging role of COTSon is extracting events (instructions, memory accesses, device requests, etc.) from SimNow without affecting execution speed. The tutorial explains how to perform a single node simulation, including supporting devices, at a desired accuracy and speed. It also covers how to use COTSon to simulate scaled up (shared-memory, multi-core nodes) and scaled out (distributed-memory, multi-node clusters) systems.
COTSon online seminar at HiPEAC
In this talk held at UPC, Daniel Ortega and Ayose Falcon provide a general description of COTSon and explain the different research challengues and solutions behind the development of the simulation infrastructure as well as some of the use cases.
Project statistics  | Analytics