Skip to content
master
Switch branches/tags
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
img
 
 
 
 
 
 
 
 
 
 
 
 

cat5tv-sbctest

Scripts to test each SBC we demo with data that can be compared from board to board.

Understanding Ģ and Ģv2

Giggles (Ģ)

Giggles (Ģ) are a cost comparison that takes cost and performance into account. While the figure itself is not a direct translation of a dollar value, it works the same way: A board with a lower Giggle value costs less for the performance. If a board has a high Giggle value, it means for its performance, it is expensive. Giggles help you determine if a board is better bang-for-the-buck, even if it has a different real-world dollar value. Total Giggle cost does not include I/O since that can be impacted by which SD card you choose. Lower Ģ is better.

Giggles v2 (Ģv2)

Like Giggles, Giggles v2 provides a value comparison. However, this number is much more accurate, resulting from LZMA CPU benchmarks provided by 7-Zip (rather than floating point tests from sysbench).

For more information about Giggles, and to compare your results to those of our pool, visit https://gigglescore.com/

Included Scripts

benchmark.sh

Usage: sudo ./benchmark.sh 79 where 79 is the USD cost to buy the board you are testing.

This script benchmarks and scores CPU, Mutex, RAM and I/O, plus runs both multi-threaded and single-threaded LZMA 7-Zip benchmarks. Using the data collected, a Giggle (Ģ) Score is generated for the board.

For the most accurate results, it is recommended to flash a vanilla Debian Base Image and then clone this repository to run the benchmarks.

mysql-benchmark.sh

Preparation: A database and user must be configured for your benchmark to run. Don't worry; you will be instructed on how to set this up on first run.

Usage: sudo ./mysql-benchmark.sh

This script will compile a temporary copy of sysbench (so tests from board to board are as accurate as possible). Then, 1 million records will be created in a test database. A total of 8 threads will be used to then benchmark the MySQL database for 1 minute, generating a report at the end.

Helper Scripts

These scripts are used by the scripts above to help with various tasks. Generally you would not run these manually.

parse.sh

Parses the output of sysbench so it can be used more effectively as data.

parsecores.sh

This script simply replies to a core switch with how many cores are on that CPU. For example, in a case of a big.LITTLE SoC, passing 0 will tell you how many cores are on the first processor, where passing the last core number (eg., 7) will tell how many cores the second processor contains.

About

Script to setup each SBC we demo with the same software.

Resources

Releases

No releases published

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •