JAICO Just in time - Articiall Intelligence - Code - Optimization
I’m a student and new here.
I’m not new to AI and have held seminars about many topics of it.
I’d like to make a suggestion to the OpenGL Specification.
First of, there’s no room for nay sayers, because they constantly fail and I only wish conctructive critics here.
I really want OpenGL to be the successor of DirectX and that it gains more technical attractivitiy
To the facts:
- Microsoft has enough ressources to buy latest technologies into DirectX and buy hardware specialized optimizations
- We have OpenCL and multiCore architecture to boost computation tasks into new level
- We have enough Computing Power to run Just in time Compilers fastly and efficiently
- We have languages that rely on Bytecode compilation technology to enable crossplatform compatibility and optimizations
- We have good evolutionary algorithms
- We have algorithms that generate evolutionary computed code that are proved to be good or better than inventend solutions
- Microcode updates increase stability in cases of errors during computatin in the cpu
- We need a large array of non-valotile gpu memory
- We have large arrays of production ready non-valotile memory called MRAM
It’s time to combine the knowledge!
I’d like my OpenGL application to analyze and benchmark my hardware to set and save settings based on my spefic hardware into the GPU MRAM (for later use of that application setings on other platforms).
And I’d like it to compute evolutionary a just-in-time bytecode compilation of core files and settings that are neeed by all applications on that platform, hardware and sofware.
I want my application to automatically track usage frequency and statistical data of OpenGL paremeters, benchmarking values and 3D/2D data for analysis by the evolutionary algorithms and more
- specific hardware optimized evolutionary generated bytecode that’s used by every application on that specific platform, operating system and software like drivers
- settings and parameters that get saved into the mram to be used on all OpenGL based applications instead of relying on the hdd/sdd
- compiling the code to bytecode to run it just in time when needed on multiple platforms and parralely on virtualized hardware
- No more relying on hardware optimizations
- No more extreme code optimizations
- No more need custom code for the graphics quality/performance and fine adjustment area, due to OpenGL’s system based analysis data on the mram
- No more need for code to utilize both the gpu and cpu as good as possible, because optimzations are evolutionary computed per application and hardware (on demand)
- Self optimizing evolutionary algorithms enhance the execution speed more and more based on application usage frequency that’s analyzed on application startup
- OpenGL Application Reporting and Statistics interface
OK, I’m sure you’ve enough inspiration now. Hope to see this idea contribution taking fruits in the code of you Core Developers.