It’s a truism that machine learning and AI requires massive amounts of compute – but CPUs and GPUs were designed for quite diffrerent applications. ThoughtWorks’ Dave Elliman gives a review of the hardware and software models needed to apply specialist operations such as cryptography, machine learning and AI, and asks what the designers of the future should be doing for us.
You can view Dave’s slides below:
Dave Elliman: Why do we need hardware for machine intelligence
You can watch Dave’s presentation below: