What Are The Potentials of AI On The Edge In Defense And Commercial Sector?

As AI is advancing, a system-of-system approach is necessary to fully explore the potentials of the technology. The system-of-system approach refers to a collection of task-oriented systems that pool their resources and capabilities together to develop a new and complex system offering more functionality and performance instead of a simple sum of the constituent systems. With the upsurge of AI on the edge and need to make intelligent decisions in such a competitive environment, the industry must examine new AI hardware and software customized to the broad range of mission across the industrial landscape.

In addition to this, continuous advancements in microprocessor and the availability of big data enables a natural foundation that powers the interest in AI-enabled devices along with ML-enhanced design and verification processes. Several AI efforts of the Defense Department have been focused on software algorithm development on existing microprocessors and hardware. Organizations within the department are deploying the potential of state-of-the-art microprocessors to create new AI-fueled warfighting capability. However, these efforts will not address the Pentagon's need for mass adaptation of AI devices. As the limitations suggest – all they depend on being tied to a computer center resource. Despite significant advances in cloud computing, there is a multitude of defense scenarios in which the resulting data latency would contribute AI-enhanced warfighting capability useless.

Development of AI-powered electronics that can learn and customize themselves as per the need and goal of the mission in a mobile environment, known as AI on the Edge, is the solution. Its implementation requires the development of optimized AI-specific semiconductors that can be configured to meet the correct mission parameters in a robust in a verified way.

Additionally, in the commercial area, application-specific integrated circuits are seeing rapid growth for edge applications that span mobile phones as well as medical, drones and industrial applications including vision and speech. Such circuits possess longer-term value proposition for edge applications because of advantages in power and decentralized independence.

Interestingly, incorporating AI on the edge does not imply the elimination of the cloud technology rather it is an implementation that yields the most efficient and optimal outcomes for the overall system. Therefore, a system design is required that leverages the computational resources available in the cloud, with high-performance, low-power system-on-a-chip (SoC) at the edge.

The size, weight and power desire for AI on the edge devices in defense and commercial sector is piloting the semiconductor industry to smaller node sizes, stretching the limits of Moore's Law (Moore's Law states that the number of transistors on a microchip doubles about every two years, though the cost of computers is halved) and creating a class of "more than Moore" customers.

The reason for this rapid growth is the state-of-the-art electronic document access tools and processes, as well as the best-in-class emulation devices that have the capability to support such development.

Moreover, to develop an AI innovation hub, the best-in-class emulation devices can be configured. This can be done while creating an AI hardware emulation center enabling the free flow of hardware design ideas and AI hardware design emulations.

Using the best-in-class emulation systems to emulate before fabrication by following commercial electronics industry design best practice, it can be assured that final AI on the hard device will accomplish first-pass success and be future-proofed.

Also, such state-of-the-art electronic document access tools and processes and the best-in-class emulation devices are now available at several Defense Department facilities.