Performance Questions?

[EvelynEve]

Hi, I'm working on some performance tests of PIXI vs WebGL. Does anyone know if those tests already exist or the best way to go about comparing the two for my usage cases?

 

Thank you in advance! 

[ivan.popelyshev]

Try to measure Windows vs Desktop Computer first.

UPD. I mean, i really dont know what tests are you talking about

UPD2. At least you have to specify the other implementation of webgl renderer except PIXI. There's no out-of-box webgl renderer in the browser, you have to write something before you actually measure it against pixi.

UPD3. Previously i thought too about tests like "canvas" vs "pixi", but then i realized that its just not possible without some kind of implementation of scene graph or just plain FOR cycle, but for webgl its not one FOR cycle, its at least 200 lines of code i think

[d13]

On July 15, 2016 at 1:08 PM, EvelynEve said:

PIXI vs WebGL.

Hi, PIXI uses a WebGL renderer under-the-hood, so in a sense it "is WebGL"

[mattstyles]

Yeah I wrote a quick 2d renderer using stackGL modules as a base, got good performance but could not match pixi in simple tests, and the code weight wasn't much lighter either (not that you'd expect an library written in a few hours to be quicker than a mature lib that has undergone many improvement and testing).

I haven't seen anything that matches Pixi for performance, extensibility and ease-of-use. Clean API, great perf, fairly tight scope, its a hard library to beat.

[ivan.popelyshev]

Dont even try it with 3d engines. We failed to incorporate gl-matrix into our code, its slower for us, somehow.

[EvelynEve]

Hi, sorry for the delay and not clarifying. We ended up figuring it out on our own! Essentially, we were trying to figure out what the FPS would be while rendering extremely large data sets. We wanted to compare Pixi to native WebGL to see what performance losses we would have while doing doing so. Note, this is using WebGL as a novice, ergo unoptimized and what basic tutorials can do (as we could expect an expert or advanced programmer or software engineer just learning WebGL to be capable of doing). Both are also only using 2D vertex sets. 

It may be interesting to you to know that we found that Pixi seems to be better optimized and corrected some intentional novice mistakes. Thus, we found that Pixi actually runs either very slightly slower (magnitude of .1 FPS) or actually a few FPS higher.

[ivan.popelyshev]

@EvelynEye Nice  I'm glad that your plots are coming good. Later, you can make your code in a plugin using https://github.com/pixijs/pixi-plugin-example . 

Btw, we have a neighbour thread where someone thinks that pixi is slow for such an "obvious" case: