# GPU accelerated final render

**URL:** <https://forums.chaos.com/t/gpu-accelerated-final-render/27455>\
**Category:** General\
**Created:** [March 8, 2011, 11:36am UTC](https://forums.chaos.com/t/gpu-accelerated-final-render/27455 "2011-03-08T11:36:58Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![RE\_FORM\_STUDIOS](https://avatars.discourse-cdn.com/v4/letter/r/59ef9b/32.png) [@RE\_FORM\_STUDIOS](https://forums.chaos.com/u/RE_FORM_STUDIOS)\
**Post date:** [March 8, 2011, 11:36am UTC](https://forums.chaos.com/t/gpu-accelerated-final-render/27455/1 "2011-03-08T11:36:58Z")

</div>

Sorry if this has been discussed before (or if it’s a stupid question), but I was curious if chaos were working on speeding up the standard vray render using GPU’s. I would have thought some aspects of the render pipeline could be sent to the gpu; ie lightcache ?  
p.

---

<div class="post-metadata">

**Author:** ![vlado](https://avatars.discourse-cdn.com/v4/letter/v/779978/32.png) [@vlado](https://forums.chaos.com/u/vlado)\
**Post date:** [March 8, 2011, 12:19pm UTC](https://forums.chaos.com/t/gpu-accelerated-final-render/27455/2 "2011-03-08T12:19:13Z")

</div>

It is not that easy; we can’t just “move the light cache” to the GPU - computing the light cache requires tracing rays, evaluating materials, lights, textures and so on. We can move some part of this to the GPU (e.g. the tracing of rays), but passing the data back and forth nearly kills any speed advantage of the GPU. Moving all the parts, including materials and lights means that we have to put essentially the entire renderer on the GPU, which is what V-Ray RT already does anyways.

Best regards,  
Vlado
