19 October 2022

Brendan Gibbs, senior vice president, automated WAN solutions, Juniper Networks

A good silicon design, that is able to scale up to meet requirements, optimises its architecture to achieve certain outcomes for customers. However, sometimes achieving one outcome comes at the expense of others. If vendors are deploying only one type of silicon, rather than different types for various use cases, key customer requirements are almost certainly going to be compromised.

This shouldn’t be the case with networks: one part of a network shouldn’t have to be compromised to improve another. Instead, networks should be run on a multi-silicon strategy, utilising platforms with different chipsets optimised for different roles in the network. In order to successfully serve customers with the best possible user experiences, the right tools must be chosen for each task.

With networking silicon in particular, any design needs to weigh logical scale against throughput. Design a highly flexible ASIC (application-specific integrated circuit) that delivers versatility with logical scale for a wide variety of complex tasks, or design one that delivers amazing throughput in bandwidth-centric roles. But optimising for both in a single design isn’t possible. Ultimately, silicon design is a multi-variable equation where logical scale and throughput are opposed variables. One can’t increase without decreasing the other.

Specialisation is key

As networks have evolved over the past two decades by supporting more diverse and demanding digital services, operators have increasingly sought specialised silicon to tackle specific roles. At the highest level, one of two capabilities for optimising network ASICs is the right tool:

• Flexible logical scale: In highly dynamic multi-service edge nodes that deliver consumer broadband (BNG), and business virtual private network (VPN) functions use cases, the most important requirements are high logical scale and flexibility. At the network edge, platforms need to support complex features for each service offering and use case, such as very large route forwarding tables, flexible tunnel encapsulation, rich quality of service (QoS), firewall security filters and the ability to attach traffic management counters to each service at the same high level of logical scale. This, in turn, drives the need for huge memory sizes that can perform complex database lookups at massive scale, speed and scope for edge role.
• High throughput: In other parts of the network, like forwarding nodes for transport aggregation and core, the script gets reversed. Core nodes don’t have to host the same diversity of network functions, as they don’t directly support logical subscribers or VPNs. But they do need to process huge amounts of traffic to keep up with growing throughput demands. As a result, core nodes need to be optimised to increase bandwidth throughput and forwarding performance with required pipeline and memory characteristics for known operations such as throughput, filtering, telemetry and sampling capabilities.

The ability to make optimisations in networking silicon is desirable so platforms optimised for the needs of specific domains at different bandwidth points and service scale points can be used. It’s only by optimising for different roles in different parts of the network that experience-first networking can be achieved.

A variety of ASICs

This can be achieved through providing choice in ASICs for different roles and customer use cases in the network, as part of a comprehensive, wider portfolio. For example, customers may require products specifically and expertly designed for high-bandwidth applications in network core and peering use cases.

A choice of ASICs offers the flexibility to deploy platforms optimised for the most demanding logical scale or throughput requirements that different parts of the network need.

Development of new silicon constitutes an enormous time and money investment.

However, investing in flexibility and choice in order to offer customers the best tools for the job without compromise, will pay off in the long-run. Ultimately, the success of a silicon strategy lies in two key words: choice; flexibility. It is these ingredients that will empower the delivery of experience-first networking.