2D Software Blitting+30MillionDoublePrecisionOnCPUOnly 512x384 15 Layers 40FPS on GeForce 7600GT with Pentium D(Pentium 4 with SSE2 and stuff 2.4GHz but really old) without NDA SDKs
Just a little trick with thread and only uploading 512x512 texture a frame.
Since we can't afford those NDA licenses if we don't have money and boy we little people can't even get an email back from graphics chips developing companies....
Better to use SDL though since there are no 3D code at all... just OpenGL is easier to setup and use with win32api than SDL so I used it.
Trick was using Thread with Sleep(1); (sleep for 1 millisecond) and only uploading 512x512 texture a frame. The number 512x512 is different if you are using cards like GeForce Titan. But you will soon find out if you are *hardware blitting* it gets really slow really fast even though graphics are shitty... I'm talking about of course without NVAPI NDA License and stuff.
I'm really sorry for the last trolling incident but got to say this. Just let's think for a monent. How does NVAPI with NDA and with OpenGL works on AMD and Intel cards? I was actually talking about that and I thought I wrote it but I think I didn't...
And people, don't use Mantle...nobody uses it everybody uses directx or opengl... don't get frustrated by those catch-phrase words of marketing department. If you get NDA(Non-disclosure agreement) license your 12FPS code will work like 2000FPS on GeForce 4200 Ti. And people, if you are going to get a NDA license, get it from Intel not AMD or NVIDIA it's way cheaper.(for now i think?)