Our Investment in EdgeQ

The 3G, 4G revolutions in the past couple of decades were centered around the mobile phone. 5G as the most advanced wireless protocol invented will extend applications beyond the mobile phone into the enterprise, industrial and even military arenas. However, hitherto, developments in the 5G chipset have really been focused on serving the telco/mobile space and they are closed proprietary systems. There is a need for a 5G hardware solution that is open and flexible to serve enterprise use cases.

Today, OEMs that want to proliferate the use of 5G in the enterprise have to buy a 5G chipset/system from the incumbent players but they are not allowed to change or modify this “monolithic black box” on their own. The toolchains to customise and add features and functionality are proprietary, meaning that while the solutions can be cohesive and compact, companies are locked in with a single vendor. On the other hand, the traditional chip providers lack the software expertise to build out the entire software stack (all the way to the application layer). Companies (who eschew the monolithic solutions) have to then stitch together a less optimal and clunky solution across hardware and software providers just to productionise.

EdgeQ’s base station-on-a-chip is pioneering a converged connectivity and AI platform based on the open RISC-V architecture. We see it as transformational because it is the world’s first 5G + WIFI + AI system on a chip (SoC).

1. Combining multiple chips in one SoC has till now been very difficult to achieve due to constraints around power and space. In customer trials, EdgeQ’s solution has been greatly superior in power usage. In fact, almost 50% of the cost in deploying base stations go into power — both in utilisation and cooling. Being able to offer 5G + WIFI in one chip (rather than two) greatly reduces cost of deployment. Not to mention the ability to apportion cores to then tackle AI use cases enabled by 5G access or 5G accessibility enabled by AI. This edgein power and cost is all the more critical as deployment scales with edge
2. It is fully customisable at the firmware (L1) layer as well as on the application stack. It essentially allows us to differentiate our base-stations/small cells with customised functionality. Think specially cyber hardened base stations or 5G capability unique to us e.g. very low latency in an autonomous warehouse robotic fleet.
3. It is software programmable; we can dynamically change the functionality on the fly much like an FPGA but with a fraction of the power needs. This is important as it allows re-programmability as 5G standards get finalised as well as the ability to repurpose chips for different use cases rapidly.
4. Being based on RISC-V means it boasts an open cellular architecture with an open-sourced toolchain. It gives us the flexibility to choose to work with different providers and vendors across the stack as the industry matures.
5. Lastly, what EdgeQ provides is an end-to-end solution — both the hardware and software (L1/L2/L3) stack. This is the only offering in the market that provides the full stack that will greatly help companies accelerate time to production.

They operate on a SaaS model, where the subscription can be tiered to different services (e.g. scaling from nominal to advanced 5G features such as ultra-reliable low latency communications, geo-location services, massive MIMO, fine-grained network slicing, as well as extending compatibility to other legacy wireless protocols).

EdgeQ is currently focusing on base stations and small cells (~$10B TAM) but can potentially take on the client device market (laptops, tablets, edge devices) which boasts a TAM many times over. The 5G market is undergoing this big shift towards more open standards, and much like how the killer applications of 3G, 4G were backward-looking (i.e. streaming only made possible after 4G invented), we believe we are really at the cusp of enabling new enterprise applications, companies and a whole new ecosystem that will be born out of a more open 5G standards.