pixel - Fotolia

Cloudian spins off edge computing subsidiary Edgematrix

Cloudian's experiments in edge computing resulted in the creation of Edgematrix, a subsidiary focused on AI use cases at the edge. The new company will use the HyperStore platform.

Object storage vendor Cloudian today launched Edgematrix, a subsidiary focused on providing AI processing at the edge.

This new company will use Cloudian's HyperStore object storage system to store and manage data, but its core product is the Edgematrix AI Box.

The Edgematrix AI Box is a ruggedized device that sits wherever cameras, sensors and other edge devices are set up to enable AI applications to process data at the source. The hardware uses Nvidia GPUs to give computers the processing power to make decisions at the edge instead of waiting for the data to be transmitted to a central hub first.

Cloudian had initially developed AI Box for some of its Japanese customers who were specifically looking for a device with enough processing power to make decisions at the edge. Use cases included cameras that analyzed year, make and model of cars to display targeted ads on a nearby billboard, automated identification and removal of defective products on an assembly line and changing traffic lights in real time to manage congestion based on current circumstances.

Edgematrix is a majority-owned subsidiary of Cloudian. It has received $9 million in Series A funding from strategic investors NTT Docomo, Inc., Shimizu Corporation and Japan Post Capital Co., Ltd. and from Cloudian CEO Michael Tso and Cloudian board director Jonathan Epstein.

"The little project had taken a life of its own," Tso said of AI Box.

Tso said he decided to spin off the AI Box technology into a new company because he wanted Cloudian to focus on its core object storage business.  Cloudian competes with Dell EMC, IBM, Scality and others in the object storage space.

There's an image angle to this -- Cloudian is seen as a data storage company, not an AI company.
Chris EvansOwner and director, Brookend Ltd.

Chris Evans, director of storage consulting firm Brookend Ltd., said the spinoff seems strategically sound because separate funding means separated risk. The market performance and development roadmap of the object storage portion of the business would not have any impact on the AI development part, and vice versa. It also allowed the fledgling company access to partner companies who are connected to the investors as well as those companies' clients. Evans said keeping Edgematrix as a separate company is also good for branding.

"There's an image angle to this -- Cloudian is seen as a data storage company, not an AI company," Evans said. "With a completely separate brand, the company [Edgematrix] can focus purely on delivering AI solutions."

Evans also said that although object storage has great potential, its adoption is hindered because customers don't know how to use the technology to solve their use cases. Edgematrix is a good example of a product that uses object storage in conjunction with AI and ML applications, and Evans believes there needs to be more like this if the technology is to take off.

screenshot and render of Edgematrix
The Edgematrixdevice is ruggedized and performs processing and data collection at the edge, enabling real-time decision-making for AI.

Tso said customers in the United States can buy AI Box through Cloudian.

Tso pointed out that demand for AI technology in Japan is notably higher than in the U.S., and it was a strategic decision to focus on Japan first. He is actively working with some U.S. and U.K. customers and discussing how to use Edgematrix's product for their use cases, but he said Japanese customers have been quicker to adopt AI products.

"Japan is an early adopter of any AI technology," Tso said. "These Japanese customers move from conversation to deployment pretty quickly."

Dig Deeper on Storage architecture and strategy

Disaster Recovery
Data Backup
Data Center
and ESG