POV-Ray : Newsgroups : povray.programming : CUDA - NVIDIA's massively parallel programming architecture : Re: CUDA - NVIDIA's massively parallel programming architecture Server Time
17 May 2024 06:00:02 EDT (-0400)
  Re: CUDA - NVIDIA's massively parallel programming architecture  
From: Warp
Date: 14 Nov 2006 12:24:30
Message: <4559fbce@news.povray.org>
PaulSh <nomail@nomail> wrote:
> Given
> that V3.7 is going to be fully threaded, what would be the possibility of a
> CUDA version?

  I would say that basically the only possibility for this would be if a
competent C++ compiler can automatically compile C++ code for it. Other than
that, hardly.

  (This is nothing new. We go back to the days of AMD's 3DNow! and Intel's
MMX and SSE: Only automatic compiler support has brought them to POV-Ray,
nothing else.)

-- 
                                                          - Warp


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.