# Is this optimization trick still true?

**URL:** <https://forums.imgtec.com/t/is-this-optimization-trick-still-true/2478>\
**Category:** PowerVR Insider\
**Created:** [July 18, 2016, 11:20am UTC](https://forums.imgtec.com/t/is-this-optimization-trick-still-true/2478 "2016-07-18T11:20:02Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![starportx](https://avatars.discourse-cdn.com/v4/letter/s/91b2a8/32.png) [@starportx](https://forums.imgtec.com/u/starportx)\
**Post date:** [July 18, 2016, 11:20am UTC](https://forums.imgtec.com/t/is-this-optimization-trick-still-true/2478/1 "2016-07-18T11:20:02Z")

</div>

from OPENGL ES GUIDE of IOS,

> **[Best Practices for Shaders](https://developer.apple.com/library/archive/documentation/3DDrawing/Conceptual/OpenGLES_ProgrammingGuide/BestPracticesforShaders/BestPracticesforShaders.html)**
>
> Describes how to use OpenGL ES to create high performance graphics in iOS and tvOS apps.

they say:  
[blockquote]Note: It may not seem obvious, but any calculation on the texture coordinates counts as a dependent texture read. For example, packing multiple sets of texture coordinates into a single varying parameter and using a swizzle command to extract the coordinates still causes a dependent texture read.[/blockquote]  
as far as I know, it’s PowerVR chips only, but is this still true to new PowerVR GPUs ?

---

<div class="post-metadata">

**Author:** ![shaun.heald](https://avatars.discourse-cdn.com/v4/letter/s/ecd19e/32.png) [@shaun.heald](https://forums.imgtec.com/u/shaun.heald)\
**Post date:** [January 13, 2017, 9:22am UTC](https://forums.imgtec.com/t/is-this-optimization-trick-still-true/2478/2 "2017-01-13T09:22:19Z")

</div>

Hi,

Dependent texture reads are significantly more efficient on PowerVR Rogue (Series 6) Graphics Cores than SGX (Series 5). However, there are still small performance gains to be had, so applications should (if possible) calculate UV coordinates before fragment shader execution.

Regards,  
Shaun