More details can be found at CFP (pdf or txt).

Download Materials

*The most recently updated files are indicated with red date color.

Contest Details

The detailed description of the ISPD 2013 contest can be found in this file: ISPD_2013_Contest_Details.pdf. The file contains the information about: (1) submission information; (2) contest evaluation; (3) benchmark files & scripts; and (4) sizer/timer interaction. Teams are advised to read this file as it has important information regarding the benchmark suite and utility scripts.

• Choice of timers: there are three options for the contestants to perform timing analysis in their frameworks:
  • The contestants can implement their own timer;
  • The  contestants call Synopsys PrimeTime^®* timing engine from their frameworks.  If option (2) is chosen,  the organizers will provide the necessary scripts and APIs to call PrimeTime®  as a black-box (with a simple function call), and the contestants have to use this infrastructure;
  • The contestants write your their PrimeTime^® script to collect timing information. The contestants can use this option at their own risk.
• It is noted that option (2) and (3) require free academic license of PrimeTime^®. The contestants are advised to contact the organizers as soon as possible if their university doesn't already have this license. Also, the organizers will send the PrimeTime^® scripts only to the contestants who register for the contest and sign an acknowledgment form. The more details about option (2) and (3) can be found in the "Sizer/Timer Interaction" section of ISPD_2013_Contest_Details.pdf.

* Other names and brands may be claimed as the property of others.

ISPD-2013 Contest Benchmark

The ISPD 2013 Discrete Gate Sizing Contest uses realistic benchmarks to evaluate all the contest submissions. All benchmarks consist of the verilog netlist, constraint file, parasitics file and cell library. Please read ISPD_2013_Contest_Details.pdf and ISPD_2013_Benchmark_Notes.pdf for benchmarks details.

Sample benchmarks and related parsers for the ISPD 2013 contest can be found here.

Final ISPD2013 benchmarks used for evaluations can be found here.

Contact Information

• Please send the emails to ispd2013contest@gmail.com.

• Please add "ISPD2013" to your subject line to get a quick response from the contest administrator!

Schedule

Submission Information

• Please try to submit a static binary as it helps with portability.

• Each team is allowed to submit a single binary that should run on all benchmarks.

• The submitted work can be either single threaded or multi-threaded version.

• In order to continue the contest, the contestants must complete at least one alpha submission during the alpha submission periods (from 01/27/2013 to 02/16/2013). Please check the evaluation metrics for details. Teams that do not submit any alpha executable by February 16, 2013, will be removed from the contest.

• No pre-computed information can be used to influence the current run. The run directory will be cleaned prior to each run.

• The officially supported programming language will be C/C++. For other languages, please check with the contest organizers first.

• The library for the contest is the standard C/C++ library. For any parallel library or third party public domain library, please check with the contest organizers first. 

• You are allowed to implement your own timer, but the final timing evaluation will be done using Synopsys PrimeTime*.

• System Specification:

  • 64 bits machine
  • #cores: 16
  • Memory: 49152Mb
  • Processor speed: 2.93GHz

* Other names and brands may be claimed as the property of others.

Evaluation and Ranking

There will be two separate rankings.

• Primary ranking: Solution quality will be the main metric. Runtime will be used for tie-breaking.

• Secondary ranking: Both solution quality and runtime will be important. Multi-core implementations are encouraged!

The details can be found in evaluation metrics slides. Final submission deadline is February 24th. Also, you must submit an alpha binary by February 16. Some changes were made from last year to this year contest, please pay attention to:

• Runtime formula was changed, runtimes were slightly reduced.
• Alpha submission testing process.
• "-fast" option for secondary metric evaluation.

Evaluations will be run on a machine with the following configuration. The evaluation program can be found here and is explained in evaluation scripts slides.

• 64 bits machine
• #cores: 16
• Memory: 49152Mb
• Processor speed: 2.93GHz

Supported Libraries

Below is the list of C/C++ libraries are allowed and supported by the contest organizers, if you wish to use a library not listed below please contact contest organizers first. Supported libraries:

• Boost (1.36.0)
• TBB
• GLPK
• OpenMP
• pThreads (2.11.1)

Call for Participation

 For the contest announcement and call for participation, please see pdf or txt.