5GC: The Core, the Container and the Bare Metal
The future of mobile switching is 5GC; future of VNFs is containers & future of container hosting and encapsulation will be bare metal.
Can 5G be called “real” unless the core is SA? Can SA be deployed meaningfully unless the VNF is based on containers? Can containers be leveraged effectively unless they are deployed on bare metal? Kaustubha Parkhi, principal analyst at Insight Research, shares the market forecast verdict for the above questions.
Does the title ring a bell? It should. After all, market forecast dissertations are known to be dour, so it a bit of zing is in order. It also helps that the discussion topic requires unraveling various layers to make sense of the technology-market force intersection.
The technology in question is 5G Core (5GC).
This article is a market forecaster’s perspective of the future of 5GC deployments. Here are the layers that we chose to unravel:
- 5GC versus evolved packet core (EPC)
- Containers versus virtual machines (VM) as the primary vehicles for virtualized cores, which include 5GC and EPC
- Bare metal versus VMs, as the encapsulation or hosting arrangement for containers in virtualized cores, principally 5GC
This article draws from the latest market research report from Insight Research – “Virtual Core – Gateway to the “Real 5G”.
5GC “Stands Alone”
All discussion surrounding 5G finally boils down to this – How “real” is the 5G implementation? As we all know by now, the reality hinges on whether the implementation has gone all out with standalone (SA) 5G Core.
SA is the real deal. Why? Because SA is that fine comb that slices the complex hotchpotch of 5G network users into fine and distinguishable strands of distinctive services that carry distinctive traffic and are governed by the distinctive quality of service parameters.
Why isn’t everyone doing it soon enough when SA is so good?
Some of the reasons are as old as time itself. Our report “Virtual Core – Gateway to the “Real 5G” lists some of the reasons:
- Geopolitical factors, read the clash of US and China, which has constricted market access for the vendors in the respective ‘blocks,’ such as Ericsson and Nokia on one side; and Huawei and ZTE on the other.
- Plain old telco ennui! When I can get my data transfer rates up by a handsome factor by just altering my RAN, why should I invest time, money and additional resources to get my core altered as well, just so that I could offer network slicing?
These factors are largely permanent while the actors keep changing their intensities and identities whenever a new technology rollout is on the anvil.
The “x-factor” causing the extraordinary delay was undoubtedly the covid19 pandemic.
Operators have now learned to coexist with covid19. In fact, a few of them have found that covid19 is good for business, registering all-time high annual revenues in their latest financial years.
Can we now say that we are ready for 5G-SA? Sure thing.
The figure below is excerpted from our report Virtual Core – Gateway to the “Real 5G”. It shows the market share progression during 2021-2026 for EPC and 5GC.
Market Share Progression for the Overall Virtualized Core; by Generation 2021-2026 (%)
Source: Insight Research
5GC is a proxy for SA technology. Expectedly, the market for 5GC will outpace the EPC market.
The next five years will witness the building of the proverbial perfect storm for these use-cases with the following ingredients:
- AI, ML and IoT constructs drive increased industrial automation
- Growing demand for private networks
- Remote management and interactions across domains and industries
- Decentralized workforce
Each of the above ingredients results in a network slicing use case. Advantage 5GC.
The upbeat mood surrounding 5G SA will thus get only deferred, not debarred!
There is, of course, the minor issue of the cost differential by EPC and 5GC, which may go up to 50% when compared to apple-to-apple. However, this issue will not be a showstopper, given the advantages that 5GC brings to the table.
See More: 5G and Its Transformative Effect on Business and Cybersecurity
Who Will Win the Virtualization Battle: Containers or VMs?
5GC will be a virtualized core, which means it will be a virtual network function (VNF). Why VNF? Because VNFs disaggregate the network function stack and allow it to run on general-purpose switching hardware. In theory, switching logic developed by a given developer should be portable across hardware platforms as long there is an interfacing layer.
Sounds familiar? It is. We are talking about virtual machines (VM) and hypervisors here. Indeed, among the VNF implementation types, the primary vehicle so far has been the VM. As VM is a mature technology, it was widely expected that the progression of VM to the telco environment would be largely smooth, given the familiarity of the industry with the technology.
But that was not to happen. There was one stumbling block – a universally acceptable sleek, and reliable orchestrator. It is commonly known as NFV orchestrator (NFVO) or management and orchestration (MANO) functionality. Finding an elegant MANO is the holy grail of VM-based VNFs. There are several open-source and proprietary MANO solutions available in the market. For a relatively nascent concept, such as NFV, however, requires an orchestrator with a comfort level that spans operators, system integrators and solution vendors.
‘That’ orchestrator curiously was found in another virtualization technology. The name of the technology is containers, and the name of the orchestrator, of course, is Kubernetes. Indeed, container-based VNFs are all poised to be the growth engine that will drive the VNF technology, including 5GC.
The figure below, excerpted from Virtual Core – Gateway to the “Real 5G,” shows the market share progression of the two virtualization technologies during 2021-2026.
Market Share Progression for the Overall Virtualized Core; by Virtualization Type 2021-2026 (%)
Source: Insight Research
While the above metric includes 5GC and EPC, it should be noted that the 5GC SA cores are predominantly container-based.
Apart from having a solid orchestrator, containers have a lot going for them. Containers are compact. Network function logic in containers resides closer to the hardware when contrasted with VMs.
And then there are microservices. Microservices literally dissect the application logic into minute functional strands. The amount of flexibility they offer to the designers cannot be matched by VM-based VNFs. Advantage network slicing. Advantage containers.
What Will Encapsulate the Containers: VMs or Bare Metal?
Let us now fine-grain the discussion on container-based core even further.
The figure below, sourced from the mentioned report, delves into an essential aspect of how containers are deployed and what the future holds for each of these deployment methodologies.
Market Share Progression for Container-based Core, by Deployment Methodology 2021-2026 (%)
Source: Insight Research
Containers encapsulated in VMs may sound counterintuitive but is actually very common. Again, the reason for that is familiar – well, it is the designers’ familiarity with VMs!
Bare metal hosting of containers leverages the proximity of the container application logic to the underlying switching fabric without the burden of intermediaries – hence the name “bare metal.” Telcos require skilled core computing hardware and software resources for designing and managing bare metal deployments. This can pose a challenge for a majority of telcos. The challenge explains the initial hesitation towards bare metal and familiarity-driven embrace of VM encapsulation.
Change is, however, clearly in the air – advantage bare metal.
The latest research forecasts that the future of mobile switching is 5GC; the future of VNFs is containers, and the future of container hosting and encapsulation will firmly belong to bare metal.
MORE ON 5G:
