Skip to main content

Accelerate AI and Vision Designs in Edge Compute

This is the third of a three-part series that describes the journey toward designing an AI-based system. The second blog covered The Journey from Prototype to Production for AI at the Edge.

As more embedded devices are installed, edge computing — or computing closer to data generation by sensors, PLCs, cameras, etc. — is taking center stage in today’s IoT era. Edge computing, which becomes necessary due to latency, security, and bandwidth/cost issues, requires efficient data processing at or near the source of data generation, which happens to be those sensors or embedded devices.

Due to diverging business/application needs, the types and locations of the sensors can vary. By that same token, the data collection and analytics processes can differ. Deep learning (or AI) can be a boon to those systems that collect massive amounts of data. But to be truly efficient, the software and hardware must be optimized with the application. All edge compute systems will be designed within a certain set of constraints, including form factor, power, cost, and/or performance. To attain maximum performance within those constraints, developers will require a set of tools that help them reach their objectives.

Development Acceleration

For IoT developers in particular, learning new tools; integrating the hardware and software; optimizing; and deploying all take valuable time, which is critical in prototype development. Hence, starting with scalable, flexible, and integrated developer kits lets them focus on their value-add and creativity. They want a seamless integrated environment where hardware and software work together to yield maximum performance. They also value the guidance, assistance, and technical support to quickly overcome barriers to development.

In short, IoT developers want (and need) a developer-centric, user-friendly, one-stop solution that leads them through their development journey of prototype to product to deployment to maintenance.

Find the Tools for Your Needs

In addition to Intel®’s strength in hardware portability, consistency, and compatibility across CPUs and accelerators, we continue to make major investments to improve the usability of our developer environment. For example, to simplify the developer experience, Intel optimized and simplified its tools portfolio, merging multiple tools into a single comprehensive suite, Intel® System Studio. Intel has also implemented consistent and long-term customer support by creating teams of Technical Consulting Engineers who train customers and help them use the software tools to optimize IoT or edge compute solutions.

In the continued theme of streamlining and centralizing, the Intel® Developer Zone, an IoT development resource that includes software tools, hardware development kits, code samples, reference implementations, starter guides, and productization and commercialization support, is the one-stop solution for all commercial developers. That solution also includes ecosystem partners — ISVs, SIs, CSPs, and edge stack providers.

Here, developers can set up the prototyping environment with a developer kit, such as the IEI Tank* AIoT Developer Kit, which comes pre-loaded with:

  • Intel® Distribution of the OpenVINO™ toolkit – To develop applications and solutions that emulate human vision using convolutional neural networks (CNNs), the toolkit extends workloads across Intel hardware (including accelerators) and maximizes performance.
  • Intel System Studio – Boost performance, power efficiency, and reliability for system and IoT device applications with this all-in-one development tool suite.
  • Arduino Create* – A cloud-based IDE that fully supports Intel libraries and Intel-based platforms. Export your IoT development from Arduino Create* to Intel System Studio for advanced capabilities.

See It Live

Embedded World 2019 attendees can view demos of these Intel hardware and software solutions, including everything mentioned above at Hall 1 Stand 338.

Additionally, a pair of technical presentations puts it all in perspective.

Presentation 1: Living on the Edge - A New Era of IoT with Intel® Developer Tools
Presented by Eric Gregory and Colleen Lay
February 27th, 3 PM
This presentation will cover Intel's new developer-enabling capabilities and how to access them at the Intel Developer Zone.

Presentation 2: Applying Cognitive Systems Engineering to Shape a Good Developer Experience of IoT Solutions
Presented by Dr. Jongsoon Park
February 27th, 5 PM
In this presentation, Dr. Park will explain how incorporating the cognitive systems engineering approach into product development can significantly improve the developer experience and have a business impact in the IoT domain.

More Resources

Bill Pearson is the Vice President, Intel Architecture, Graphics and Software and the Director, Internet of Things Developer Programs at Intel Corp. He is responsible for enabling developers and applications on Intel’s IoT products. He and his team deliver developer assets, and engage developers directly to drive strategic application enabling, architecture conversions, porting and optimization supporting the Intel IoT product portfolio. Bill has held a variety of technical and management roles during his 22 years at Intel.

*OpenVINO and the OpenVINO logo are trademarks of Intel Corporation in the U.S. and/or other countries.