At the conference AMD’s Graphics CTO Eric Demers described how the company plans to evolve the GPU and have it break free from the “shackles of fixed-function designs” by heavily integrating future general-purpose processors and graphics processors, hence the titular namesake of AMD’s “Fusion” line of APUs.
Demers said that this evolution will be gradual, and with each rollout of new APUs AMD will move incrementally closer to this goal with the first set of new features appearing later this year. Some of the milestones that will be reached between now and the end of the evolution in 2014 are:
- GPU Support for C++ and other high level constructs.
- Virtual address space.
- Support for page faults.
- Memory coherence at the L2 level to allow for seamless integration and the ability to reload the same device state.
Phil Rogers, a corporate fellow at AMD said that “The first APUs from AMD dramatically increase processing performance while consuming less power and now we are building upon that achievement with our next generation of products. Future innovations are intended to make the different processor cores more transparent to programmers. They can then seamlessly tap into the gigaflops of power-efficient performance available on the APU and design even faster, more visually stunning applications on a wide range of form factors,”
Is this one of the more dramatic moves to heterogeneous computing from AMD? What do you think?