Local "Dataset" API & Hardware Cache

› Forums › Personal Topics › Unbidden Thoughts › Local "Dataset" API & Hardware Cache

This topic contains 4 replies, has 1 voice, and was last updated by  josh March 24, 2021 at 8:28 pm.

  • Author
    Posts
  • November 10, 2020 at 11:41 pm #71099

    josh

    Think of a local IT manager’s POV – if they can increase their ability to seamlessly use multi-cloud while having a local box they can keeping adding to meaningfully boost performance if relevant, then that’s probably something they like.

    • November 10, 2020 at 11:45 pm #71100

      josh

      I’m looking at Ignite there as a successful algorithm which could be implemented in some combo of system software/firmware/switching/applicatoin rather than a hosting of the Java lib.

      • November 10, 2020 at 11:50 pm #71101

        josh

        High performance Java interface to ALL YOUR FAVORITE ML packages would be popular, for instance.

  • March 24, 2021 at 8:28 pm #87787

    josh

    The OpenStack Congress was developed mostly by VMWare engineers & didn’t take off , presumably losing out to Kubernetes & other alternatives. It seems like the main application is putting brakes on undesirable cloud usage rather than selling more raw hours. The expressed goals of the project were to promote declarative policy from the users POV. There might be some elements which represent useful abstractions & implementation ideas for the LAN to cloud gateway.

  • Author
    Posts

You must be logged in to reply to this topic.