Hendrik Proosa wrote:Rajiv Mudgal wrote:On the other side, the vfx side of the post production its wait and watch..
I also use fusion, boris fx, mocha pro, octane/blender and am not seeing the same performance, but that may change soon.
Jules Urbach of OTOY (Octane render) has announced that an M1 version of Octane is in development and that is it "FAST!", so far so good.
And that 16 core neural engine is currently under utilized, so let's see.
I doubt it will make a dent in vfx, because power efficiency isn't an objective there. Efficiency is ofcourse good, more efficient at same compute power is better, but less compute for less power drain isn't a selling point. The laws of rendering are on the side of "do whatever it takes to make it go faster". Future might be ARM or some other RISC based processor family, but it most probably isn't by Apple, because Apple does not operate in a lot of areas and I think has no interest in doing that either.
It all kind of reminds me the quote by Master Zap from MentalRay who said that "ofcourse Renderman is fast, until you actually trace a ray". Ironically Mental Delay is now pretty much dead and Renderman traces rays like there is no tomorrow. Not sure if Apple is the Mental Ray here, or Renderman. But Renderman isn't sitting on its throne very comfortably, after Arnold kicked the door open with raw energy. And they are both looking over their shoulder where near-realtime gpu rasterization and raytracing are slamming left and right. But again, laws of rendering tell that near-realtime solutions will not kill off everything else.
I have a similar feeling about VFX, but imagine using 25 macmini running in slave mode and running it using home power outlet using just a 1 ton ac to cool it, its still cheaper than what I pay to external render farm. But there is a big BUT.
https://www.imdb.com/name/nm8500260/