Collect performance data by running 100 steps: first just the graphics, then
just the model, and finally the model and graphics together.
First row contains averaged values from all runs.
Note: Some identifying information will be sent along with your benchmark data. To see more
In order to measure performance over time, we save a semi-unique "fingerprint" of
each browser that runs the benchmarks. This allows us to look at performance over
time of models running in a single user's browser.
To do this, we log a number of identifying pieces on information about your
browser, such as the browser's userAgent, language settings, colorDepth, etc. This
information is then hashed into a single number.
The unique fingerprint for your browser is
It may be possible to use this information to uniquely identify your browser
elsewhere on the web. If you prefer that this information is not submitted, please
do not hit the Submit button.