Sunlight.io: Efficient virtualization key to managing a secure edge for QSR’s, retail, manufacturing and beyond

In practice, manageability is an issue for many developers. For example, while enterprises are increasingly turning to Kubernetes for edge AI and ML workloads, recent research suggested that only 42% of organizations said all applications they had developed on Kubernetes made it to “Day 2” environments.

How Sunlight is addressing the edge application challenge

Last year, Sunlight.io upgraded its Sunlight NexVisor and Infrastructure Manager software to facilitate the management of large, distributed edge deployments. The revisions enable applications of different sizes to be deployed on both x86 and ARM-based systems. While Sunlight’s technology can be used to power centralized cloud services, it is perhaps most uniquely suited for deploying at scale in on-premises configurations to run applications in retail stores or on factory floors.

Edge Industry Review recently spoke to Julian Chesterfield, co-founder and CEO of Sunlight, along with Kosten Metreweli, Sunlight’s co-founder and chief marketing officer, to talk about trends in the edge computing market, and the challenges of developing and managing applications for demanding environments.

The following comments from the interview have been edited for length and clarity.

EdgeIR: Can you give us some background on your company and Sunlight’s perspective on how the edge computing market is developing in 2022?

Julian Chesterfield (JC): The important thing that triggered the creation of Sunlight was the work that we were doing with ARM technologies in Cambridge. Our contribution to that project was building a lightweight virtualization layer that could run effectively and help to create a very small footprint, lightweight, clustered hypervisor technology. The Sunlight idea and architecture really sort of spawned out of that.

What has been very fortuitous for us is that the whole edge ecosystem has started to become a reality. It’s pretty much maturing and what we’re seeing now is the mainstream adoption of edge technologies. Sunlight is ideally positioned to be able to support that.

EdgeIR: What are some key trends you are seeing with enterprise deployments? What are some of the key market verticals that are adopting edge computing, from your perspective?

JC: For the whole edge to core ecosystem, what we’re seeing now is this requirement for enterprises to be able to start to disaggregate a lot of their compute infrastructure. Sunlight is designed to be able to support that type of usage model.

I think there are a number of markets that are very appropriate for Sunlight, but the ones that we’re seeing most traction with at the moment are around quick-service restaurants (QSRs), retail and manufacturing. These are organizations that have large numbers of locations where they need to be able to support and manage services.

In the QSR and retail environment, one of the main drivers we’re seeing is the adoption of machine learning and video analytics to be able to provide a variety of different applications and services — using technology to bring a lot of value, to make restaurants more efficient and help with more customers or reduce waste, for example.

EdgeIR: What do you see happening from the hardware and application layer in QSRs, for instance? Are QSRs moving from Windows and Intel-based platforms to ARM, for example?

JC: What we are seeing is very much what happened in the cloud. There’s a consolidation of services onto a common platform. The next step is going to be the containerization of services. And then [companies] will be able to adopt the different types of hardware platforms as well [such as] more efficient, high-performance ARM-based platforms, like the NVIDIA Jetson platform. We’re starting to see the mainstream hardware vendors adopting NVIDIA system-on-chip nodes as a core part of their offering.

Kosten Metreweli (KM): Organizations have a whole mix of infrastructure that’s now beginning to roll out in their restaurants or retail stores or manufacturing facilities, and so they need a way to consolidate that down into a more efficient, reliable, and manageable system. And now with the advances of AI/ML and CI/CD development models, you’re changing infrastructure or the applications daily. You’re increasing the cadence of changes to applications and the need to consolidate it all down into something manageable and efficient.

EdgeIR: Everyone has had challenges in managing risks on the cloud. Will containerization and edge help or just make the problem bigger? What role do you Sunlight think you can play in securing edge applications?

JC: I think there are advantages for distributed infrastructure. For instance, from a data sovereignty point of view, edge is a big deal. Perhaps in the US, the regulations I would say are still developing. In Europe, it’s always been a much bigger deal to have tighter controls around the locality of data, particularly where it involves personalized data, whether that’s credit card information or user information about shopping habits. I think enterprises increasingly have a responsibility to ensure that data locality and data sovereignty are really on a given device. The technology solutions that we provide can help those enterprises with those sorts of challenges.

EdgeIR: What are some of your key initiatives to drive growth in 2022?

JC: The managed services provider (MSP) space is an industry vertical that is starting to innovate quite fast. They have traditionally provided infrastructure in data centers for their enterprise customers and managed that remotely. Managing infrastructure at the far edge — even further outside the data center infrastructure — for things like video analytics and machine learning applications are often driving MSP interest in our technology. The MSP space is certainly a strong market for us and is an area where we’re growing and acquiring customers.

Article Topics

application development  |  ARM  |  device management  |  edge orchestration  |  HCI  |  hyperconverged  |  manufacturing  |  QSR  |  retail  |  Sunlight.io  |  virtualization