1 Comment

  • I wrote neural simulator in C++ and it's really good. I want to use it as the basis for the AI in an online, 3D first-person-shooter game but the fluid dynamics are impossibly slow.. That kind of processing is realistically impossible due to the serial nature of CPU processing.

    So I began learning FPGA and discovered it isn't difficult. In fact, being inherently parallel, it greatly simplifies fluid dynamics. This caused me to realize that conventional computers must die for the sake of human progress.

    Perhaps GPUs are the way to make that conversion happen. Not to FPGA, per se, but to fully parallel processing.

Comments have been disabled for this content.