Edge Computing’s Need for NVMe Computational Storage

Apr 23 2019

Scott Shadley

Before we explore how edge computing can benefit from computational storage, it is worth defining what edge computing is. Edge computing is an approach that places computational resources close to where data is generated. Edge computing is most often discussed in the context of the Internet of Things (IoT), and for good reason. IDC has stated that 25% of data that will be generated in 2025 will be real-time data, and 95% of that will be generated by IoT devices. Shipping all of that data from the source to a data center or a cloud is usually prohibitive, which is where edge computing comes into the picture. Edge computing accomplishes two tasks in this context: i) it provides the potential for real-time analysis of data; and ii) it significantly reduces the amount of data that has to be shipped upstream.

So how does computational storage fit into the edge computing? In addition to requiring both computing and storage resources, edge computing also presents additional challenges for conventional data processing architectures; these include limited space, power, and cooling. Just think about putting a standard 1U “pizza box” server into a self-driving vehicle or into an automated machine in an industrial environment – challenging would be an understatement. In contrast, computational storage solutions such as the NGD Systems Newport U.2 solid-state drive (SSD) packs 16TB of data storage, a multi-core ARM processor, and dedicated acceleration hardware for AI, encryption, and other capabilities into a 2.5”SSD that is only 15mm thick, and consumes less than 10 watts on average.

TECHunplugged recently wrote an independent analysis on computational storage. It includes an exploration of use cases for computational storage in this market, as well as an overview of the vendors currently building products for this market. You can either download the paper by going to the “bit.ly” listing in the illustration, or by downloading it here. It has some great points about why computational storage makes sense for edge computing. Enjoy

NGD Processor chip.

355 Goddard, Suite 200
Irvine, CA 92618

For Sales Inquiries:


For General Inquiries: