With the advent of heterogeneous computing systems that combine both main CPUs and connected processors that can ingest and process tremendous amounts of data and run complex algorithms, artificial intelligence (AI) technologies are beginning to take hold in a variety of industries. Massive datasets can now be used to drive innovation in industries such as autonomous driving systems, controlling power grids and combining data to arrive at a profitable decision, for example. This paper describes how AI can now be used in various industries using the latest hardware and software.
When the cost and complexity of a set of new technologies decreases, more opportunities arise for this new technology to be accepted and brought into mainstream business processes. More easily accessed applications combined with the increase in performance per dollar has opened up new opportunities for using AI.
New hardware is allowing for the solving of very complex problems in a fraction of the time that was previously spent performing the same task. Or, if a constant time to completion is desired (that is within the limits set by the workflows), more data or more complex applications can be used to arrive at a more precise understanding of the data to arrive at more confidence in the final understanding or action. With the explosion of data that is available for example through social media, a more complete understanding of consumer sentiment can be analyzed and more accurate decisions can be made based on consumer feedback. Limits placed on the amount of data to use can be expanded exponentially compared to previous generation systems.
Machine learning, which is a key component of an AI solution, is becoming more available to more industries through the development of new easier to use software combined with the massive parallel hardware to run these applications. Many processes that were previously performed by humans with a deep knowledge of an industry are becoming more available to a wider mass as these algorithms with massive amounts of data are commercialized. Simulations of these decisions can now be performed to understand the benefits and risks associated with these automated decisions.
The latest generation of CPUs and associated processors allow for scaling of these algorithms to thousands of systems and potentially hundreds of thousands of threads. When planning for a new environment that can assist with business decisions, it is important to understand the industry domain, the workflow requirements, the software to be used and the underlying systems that will be used for the computation. With the best-in-class choices, an innovative AI environment can be planned, setup and deployed which will achieve the desired business benefits.