Menu
Forums
New posts
Search forums
What's new
New posts
New profile posts
Latest activity
Vault
Time Travel Schematics
T.E.C. Time Archive
The Why Files
Have You Seen...?
Chronovisor
TimeTravelForum.tk
TimeTravelForum.net
ParanormalNetwork.net
Paranormalis.com
ConspiracyCafe.net
Streams
Live streams
Featured streams
Multi-Viewer
Members
Current visitors
New profile posts
Search profile posts
Log in
Register
What's new
Search
Search
Search titles only
By:
New posts
Search forums
Menu
Log in
Register
Navigation
Install the app
Install
More options
Contact us
Close Menu
Forums
Time Travel Forum
Time Machines & Experiments
Assembly Language
JavaScript is disabled. For a better experience, please enable JavaScript in your browser before proceeding.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an
alternative browser
.
Reply to thread
Message
<blockquote data-quote="Beholder" data-source="post: 244584" data-attributes="member: 14640"><p>I can show you a math trick for deriving polynomials. Store coefficients using for a + bx + cx²/2... in registers and use name aliasing for referring to input and output. Shifting your perception of which register holds which multiplicity with pre-multiplied units optimized for physical simulation will do the operation at zero cost, in other words infinity faster in that specific part by being flexible with how you define units.</p><p></p><p>Changing units is also used to sample textures efficiently using fixed precision integers so that multiple bit shifts are merged into one, unzip instructions for division and modulo, et cetera.</p><p></p><p>If you can throw in a hundred of these tricks into the same code, you may soon have a 200x performance increase over the entire application. Can't refer to my open source optimizations anonymously, but this is used a lot in image analysis and software rendering.</p></blockquote><p></p>
[QUOTE="Beholder, post: 244584, member: 14640"] I can show you a math trick for deriving polynomials. Store coefficients using for a + bx + cx²/2... in registers and use name aliasing for referring to input and output. Shifting your perception of which register holds which multiplicity with pre-multiplied units optimized for physical simulation will do the operation at zero cost, in other words infinity faster in that specific part by being flexible with how you define units. Changing units is also used to sample textures efficiently using fixed precision integers so that multiple bit shifts are merged into one, unzip instructions for division and modulo, et cetera. If you can throw in a hundred of these tricks into the same code, you may soon have a 200x performance increase over the entire application. Can't refer to my open source optimizations anonymously, but this is used a lot in image analysis and software rendering. [/QUOTE]
Insert quotes…
Verification
Post reply
Forums
Time Travel Forum
Time Machines & Experiments
Assembly Language
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
Accept
Learn more…
Top