At Kaloom™, our mission is to reduce costs per connected device, per gigabit of traffic and labor costs by simplifying complex distributed edge networks. We have been at the forefront of edge computing innovation since launching Unified Edge Fabric™, a distributed edge data center solution envisioned and developed in collaboration with Red Hat in May 2020. As service providers join the 5G ecosystem, they struggle with dynamic data flows around their conventional network siloes, resulting in an exorbitant total cost of ownership (TCO) of edge infrastructure. These difficulties are caused by a lack of programmability at the data plane and the heightened data processing requirements of 5G applications. The transformation of distributed 5G edge infrastructure through open networking becomes increasingly important as the bar is raised for these requirements.
Unlocking the economics of the distributed 5G edge
5G and the emerging applications, such as connected cars, industrial robotics, Augmented Reality, Virtual Reality, or interactive gaming, raise the bar for cost reduction, throughput, and latency that simply couldn't be addressed with the existing infrastructure. Most vendors have been implementing 5G without changing the network technologies.
This means the cost points haven’t changed either. Operators must transform their distributed 5G edge infrastructure to achieve the ultra-low latencies and terabytes of throughput required by data-rich, business-critical applications. But how can they reduce TCO while ensuring efficient resource use in the face of space and power constraints at the edge of their networks?
Our case study “Total Cost of Ownership Study for Distributed 5G/Edge Deployments” addresses this question with inputs from customers for a central office use case, showcasing how our fully programmable, automated Unified Edge Fabric™ solution helps operators navigate telecom’s shifting economics. Our Unified Edge Fabric unlocks the power of the distributed 5G edge through heterogeneous hardware that enables P4-driven data plane programmability, disaggregation and more to greatly reduce TCO.
Challenges for operators at the 5G edge
The distributed 5G edge presents challenges to operators, including increased network switch cost with the growth of dynamic traffic, and the inadequacies of traditional fabrics and current static forwarding mechanisms. The packet processing capability of proprietary hardware is too inflexible and costly to cope with 5G’s heightened compute demands. Traffic redundancy of conventional infrastructure results in congestion as compute-intensive tasks, such as load balancing, intrusion detection and more, are directed to servers elsewhere. This redundant movement of data negatively impacts operational costs and network latency.
Operators need automated software solutions that enhance control and programmability of business-critical applications by supporting dynamic data traffic growth. By leveraging the P4 language for data plane programmability, operators can improve control over their distributed 5G edge networks and enhance quality of service through forwarding rules for physical resource allocation applied at the network underlay. This is where our Unified Edge Fabric solution comes into play, allowing operators to transform their underlying edge infrastructure and consolidate network functions through shared resources.
Our Unified Edge Fabric: A programmable solution for TCO reduction
With our cloud-native Unified Edge Fabric solution, operators can enhance data plane programmability and disaggregate hardware from software, greatly reducing operational costs and eliminating resource use inefficiencies. This disaggregation allows operators to reprogram hardware without the need for redesigning or swapping. Operators can also rely less on proprietary implementation by leveraging granular telemetry and operations on commodity hardware. Furthermore, operators can achieve multi-tenancy and edge monetization through 5G network slicing and other capabilities. This reduces operational costs by up to 60 percent while ensuring their distributed 5G edge networks evolve according to the compute demands of new services they hope to bring to customers.
Unified Edge Fabric features outlined in our case study include the following and more:
- Granular optimization of space and power
- Network consolidation and hardware acceleration according to workload’s needs
- Cloud-native containerization
Unified Edge: TCO Case Study
Our case study showcases the TCO reduction operators can expect with the implementation of our Unified Edge Fabric. Our Unified Edge TCO calculator uses customer inputs to quantify cost reduction.
We substantially reduce TCO through an open ecosystem approach by integrating our fabric with Red Hat’s OpenShift container orchestration platform. This case study proves how an automated, fully programmable network fabric can replace conventional edge infrastructure, including Top of Rack switches, fabric overlay servers, Kubernetes master servers and UPF servers.
Our Unified Edge solution marks an evolution in edge computing, empowering operators with the ultra-low latency benefits of heterogeneous infrastructure at the distributed 5G edge. In addition to TCO reduction, operators can expect the following benefits and more:
- Heterogenous hardware comprised of control and data planes optimized for network functions
- Transformation from virtualized to containerized infrastructure
- A single platform for multi-vendor networks and cloud connectivity