The Future State of Disruptive Computing – (Distributed, Deep Machine learning, Embedded intelligence)

The future clearly appears to be built around hardware systems that are mostly composed of cell phone components + software + internet connectivity + data-enabled intelligence automation. The convergence of these is unbelievably powerful. We already see the rise of embedded intelligence built into hardware systems around us?—?and this is set to grow exponentially.

What’s fueling this new age of computing, but not getting deserved attention in the public sphere, is GPU (Graphics Processing Unit) computing. NVIDIA clearly stands to benefit as a early mover but others may catch up.

GPU’s are powering the rapid growth of deep learning and AI. Companies that pursue hard tech now hear about GPUs as often as we used to hear about AWS a few years ago. In fact, one could argue that deep learning could likely become the future of ALL computation.

Between on-device AI & in-the-cloud AI, all kinds of hardware systems are becoming intelligent. They can see, learn, find their own way. In fact, the use of GPU compute + Machine Learning/AI is now almost ubiquitous in modern hardware startups. From self-driving cars (stealth) & drones to home automation (stealth), physical security systems, medical devices and AR/VR. In fact, new GPU based infrastructure is literally helping the blind see and physicians make smart decisions

As cost of sensors, connectivity and compute approach zero, the question is what kind of future can be envisioned a few years out? It is easy to visualize a world where smart sensors were embedded in every dumb surface we see around us, image and voice sensors seamlessly analyzing spaces and activity for us, and our decisions always supported by tons of data. It can also be private, where AI would make us feel our 6th sense was actually real.