Architecture Of Data In The IoT

The challenge is to balance low cost and low power against the demand for high performance for devices connected to the Internet of Things—and to make them secure.


It’s clear that the Internet of Things (IoT) is no longer just a visionary concept. It’s on the verge of becoming a new reality. But this new reality is heavily dependent on the ability of an infrastructure to capture, secure and move data across networks.

Given the tremendous amount of data that already surrounds us, this emerging paradigm will enable us to not just track and measure, but to analyze, predict and eventually use data to autonomously take action. Billions of devices connected all over the world increasingly are capable of using data to achieve countless tasks. From the common task of geo-location, to more advanced applications like automobile accident prediction and avoidance, we are just starting to scratch the surface of what is possible.

To enable these new systems, designers have to balance between the demand for high performance against the need for low cost and low power. This is true for both client devices, where there is a seemingly endless number of new IoT products, and also in data centers, where memory density, security, power and cost are all essential to support growing data and performance requirements. As we progress toward the reality of a truly interconnected world, designers are taking a more holistic view of the IoT ecosystem.

Capture. Objects and machines within IoT ecosystems need to be designed to gather information and quickly act upon relevant data. Ranging from collision detection to occupancy sensing to facial and gesture recognition, smart sensors are the fundamental components of enabling the IoT. However, in order to be implemented properly, they must be designed to be very low cost and operate with the lowest possible power consumption.

Secure. As more sensitive and valuable data is captured and communicated across interconnected devices, robust security becomes paramount. Simple, but effective, security embedded at the hardware-level, combined with flexible digital asset management services, transforms the Internet of Things into the Internet of Trusted Things.

Move. The method through which data is processed and communicated is a foundational element to a functional infrastructure that can scale to support the more than 30 billion connected devices projected to be a part of the IoT ecosystem by 20201. Specifically, memory interfaces, and the corresponding new memory architectures are vital to optimizing power consumption and managing the trade-offs of between local versus cloud-based data analysis and processing. The competing goals of lowering power and achieving higher performance are seen in both mobile clients and datacenter hardware. Without dramatic innovations in battery, energy and density, the mobile client performance of today suffers when confronted with tightening power constraints affecting subsystems. Power challenges arise from the sheer scale of aggregating tens of thousands (or more) of servers into one facility. Moving data from one place to another — whether across a chip, between chips in the same system, or longer distances between servers and racks of servers — consumes a large fraction of the power budget in mobile and datacenter systems.

As the value and importance of data grows, continued innovation is needed to develop solutions that address emerging requirements for how data is captured, secured, and moved within and between modern connected clients and datacenters. We invite you to continue reading this blog as we further explore these and other topics over the coming months.