Logo by AGUS - Contribute your own Logo!

END OF AN ERA, FRACTALFORUMS.COM IS CONTINUED ON FRACTALFORUMS.ORG

it was a great time but no longer maintainable by c.Kleinhuis contact him for any data retrieval,
thanks and see you perhaps in 10 years again

this forum will stay online for reference
News: Visit the official fractalforums.com Youtube Channel
 
*
Welcome, Guest. Please login or register. September 19, 2020, 11:20:53 AM


Login with username, password and session length


The All New FractalForums is now in Public Beta Testing! Visit FractalForums.org and check it out!


Pages: [1]   Go Down
  Print  
Share this topic on DiggShare this topic on FacebookShare this topic on GoogleShare this topic on RedditShare this topic on StumbleUponShare this topic on Twitter
Author Topic: anyone playing with AVX instructions yet?  (Read 1009 times)
0 Members and 1 Guest are viewing this topic.
cbuchner1
Fractal Phenom
******
Posts: 443


« on: January 26, 2012, 12:10:13 PM »

Hi,

has anyone tried doing some fractals with the new Intel AVX instruction set yet?

Scalar programming is so 20th century... wink
« Last Edit: January 26, 2012, 05:03:57 PM by cbuchner1 » Logged
cbuchner1
Fractal Phenom
******
Posts: 443


« Reply #1 on: January 26, 2012, 12:42:19 PM »

oh, by the way what's really irking me about the SSE2 and AVX instruction sets is that transcendentals such as exp, log, sin, cos aren't available in hardware. You always have to implement them explicitly - much unlike CUDA or OpenCL where the GPU hardware will give you the result in a single instruction.

I remember seeing some Mandelbulb code that avoided transcendentals as much as possible (losing the feature of allowing variable exponents though), so this method is probably the way to go for a fast vectorized implementation.

« Last Edit: January 26, 2012, 12:59:03 PM by cbuchner1 » Logged
DarkBeam
Global Moderator
Fractal Senior
******
Posts: 2512


Fragments of the fractal -like the tip of it


« Reply #2 on: January 26, 2012, 05:13:03 PM »

oh, by the way what's really irking me about the SSE2 and AVX instruction sets is that transcendentals such as exp, log, sin, cos aren't available in hardware. You always have to implement them explicitly - much unlike CUDA or OpenCL where the GPU hardware will give you the result in a single instruction.

I remember seeing some Mandelbulb code that avoided transcendentals as much as possible (losing the feature of allowing variable exponents though), so this method is probably the way to go for a fast vectorized implementation.



You can always use the good old floating point routines. I am very familiar about those sleeping
Logged

No sweat, guardian of wisdom!
Pages: [1]   Go Down
  Print  
 
Jump to:  

Related Topics
Subject Started by Replies Views Last post
Develop a programme based on drawing instructions Let's collaborate on something! raphuu 3 1207 Last post June 09, 2010, 08:47:54 PM
by raphuu
Playing with Buddhabrots Images Showcase (Rate My Fractal) Fractal Ken 5 592 Last post November 23, 2012, 03:25:26 AM
by Fractal Ken
Playing catch-up in the Finger Lakes Region of NY. Meet & Greet willclark218 4 602 Last post January 18, 2013, 10:47:11 PM
by willclark218
playing billiard on relativity table Mandelbulb3D Gallery abbaszargar 0 496 Last post February 22, 2013, 02:06:30 PM
by abbaszargar
Iterated tie instructions Fractal Humor youhn 0 659 Last post June 03, 2014, 09:39:35 PM
by youhn

Powered by MySQL Powered by PHP Powered by SMF 1.1.21 | SMF © 2015, Simple Machines

Valid XHTML 1.0! Valid CSS! Dilber MC Theme by HarzeM
Page created in 0.137 seconds with 27 queries. (Pretty URLs adds 0.005s, 2q)