設計工具
company

Industrial quotient: Living the industrial AI mindset

Micron Technology | August 2023

Today, there are over 15 billion connected internet of things (IoT) devices. And you can expect twice as many — over 29 billion — by 2030. Big machines using big data are transforming the industrial market, ushering in the next industrial revolution and relying on complex AI workloads to manage the rapidly growing sensor data.

Yet, today’s smart manufacturing isn’t something new — rather, it’s building on the shoulders of the previous revolutions: mechanization in 1780, electrification in 1870, automation in 1970, and connection in 2011. Through every stage, industry isn’t changing, it’s evolving, adding new requirements to become even smarter, even faster.

To that end, today’s AI-powered industrial automation has more requirements than ever before. At the center of this transformation are memory and storage — and harsh industrial environments demand specific characteristics of performance. These include lifecycle support, ruggedness, reliability, superior quality and product enhancements for specific application requirements.

Industrial AI increases the demand for memory and storage

We see this across industries, including medical applications, where advanced imaging is driving GPU-class machines with high-bandwidth memory (HBM), high density DDR5 DRAM and multi-terabyte SSD storage. In video security, edge storage systems need to store and process high-resolution video, and then run AI-powered analytics on the video data. This is very demanding on resources, as AI models require a lot of memory and storage space.

orange and black horizontal chart showing that cameras need more memory and storage

Perhaps nowhere is the requirement for memory and storage changing more than in industrial automation, where industrial IoT edge servers drive two to three times cache and code storage and four to five times data storage versus a standard server to handle the demands of AI workloads at the edge.

orange and black horizontal chart showing that industrial cameras need more memory and storage

Industrial transformation requires new edge devices and gateways to convert data into insights. From self-driving cars to aerospace, from video security to the shop floor, our industrial memory and storage solutions deliver the required performance, latency and security — from infrastructure to AI at the edge.

In fact, at Micron, we leverage data analytics and AI in our own manufacturing processes. Silicon manufacturing is an extremely complex process, taking months and involving some 1,500 steps. Micron employs sophisticated AI in every step of this process, dramatically improving accuracy and productivity. The benefits are many, including higher yields, a safer working environment, improved efficiencies, and a sustainable business.

Of course, none of these components works without memory. Memory, in fact, puts the “intelligence” in AI, providing it with data to run its algorithms, and context for its actions and reactions.

Developing an industrial AI mindset

When choosing a supplier for the industrial AI space, it is important to keep in mind that standards should be higher than those for a typical consumer application. Things like longevity, ruggedness and reliability are paramount when deploying solutions in harsh environments or other areas with a longer upgrade path.

At Micron, we agree with this philosophy and have developed a set of core values by which we engineer our products and operation models. Our guiding principle is a powerful framework for today’s industrial applications we call the Micron “Industrial Quotient” or IQ. We have partnered with many other like-minded thinkers in the industry to develop these five IQ characteristics:

  • Longevity: Micron supports a five-plus year product lifecycle. Industrial applications are built to last and as such require extended product support.
  • Ruggedness: Improved performance in extreme environments, including temperature, shock, corrosion, humidity and radiation.
  • Reliability: Performance stability is measured with key benchmarks around annual failure rate, failure in time, and mean time between failure endurance.
  • High Quality: Ongoing, extensive testing on all processes minimizes variance and maintains consistency.
  • Application Optimization: Our product enhancements (hardware and firmware) meet app-specific requirements, including auto-scan/auto-refresh, firmware health monitoring and API encryption.

At Micron, we’ve delivered leadership and innovation in the industrial market for over 25 years. Our “industrial mindset” means we deeply understand these use cases. Our holistic focus on quality, reliability and ruggedization ranges across our wide portfolio of application-specific solutions. And it’s all backed by our industry leading fab longevity support, where we have made significant investments.

Shared values

The core values we carry to support industrial AI applications are not unique to Micron, however. Many of our industrial customers share the same mindset and values. The rugged and reliable memory and storage solutions we produce are components of machines that carry the same standards of ruggedness and reliability. In Micron’s leading smart manufacturing facilities, we use machines built by our customers that are enabled with our own memory and storage solutions because we know it is a solution we can trust. For example, the AI-driven machine vision tools we use to detect and correct manufacturing deviations in our products are enabled with the very memory and storage components the machines are helping to create.

In industrial IoT, this level of core process knowledge is critical. At Micron, our rich heritage and deep understanding of the AI industrial IoT ecosystem ensures we get you to market — with unparalleled total cost of ownership.

This ongoing dedication to these core values has made us the undisputed leader in industrial memory and storage, delivering design innovation and differentiation backed by our worldwide system labs and engineering, sales and distribution networks. In the industrial IoT space, companies should make sure to invest time researching solution options and choosing those firms who share the same set of values and can get the company to market with the right solutions.

The innovations that our people create fuel today’s industrial revolution, enabling advances in AI applications. After all, we supply the critical components to the creators making AI, collaborative robots, computer vision and other cutting-edge applications. When they need innovation, they come to Micron.