Virtual RAN still seems to be not worth the effortVirtual RAN still seems to be not worth the effort

Laboratory speed tests are to 5G what a drag race in a sports car is to the average motorist. Just as no driver of a Nissan Note on London roads will ever have need for a 200mph engine, no smartphone user requires a 5.5Gbit/s mobile service to watch Netflix, play games, send messages or visit Elon Musk's social-media house of horrors (or digital town square, as he calls it). In any case, most of this probably happens on Wi-Fi rather than cellular networks.

But none of that stops technology developers from pushing the envelope, and Verizon, one of the biggest telcos in North America, now claims to have "shattered" the 5.5Gbit/s "speed barrier" in a lab trial. As if to prove how unnecessary it all is, Verizon in its press release included the real-world detail that 5.5 Gbit/s is enough to download 266 Taylor Swift albums in a minute. Because that's exactly the sort of thing a normal person wants to do.

Still, in a typically quiet week of the year for telecom news, some of the technology specifics hold interest. The trial used Samsung's 5G network equipment and a device powered by a Mediatek processor. Verizon also appears to have combined spectrum from six different channels of sub-6GHz spectrum – a process known as carrier aggregation – to spin the needle on that speedometer.

It's a laboratory showcase, too, for the capabilities of general-purpose technology, with Samsung stumping up its virtual radio access network (RAN) smarts in the trial. This naturally meant running 5G RAN software on a general-purpose processor supplied by Intel, rather than a chip tailor-made for those 5G functions.

Samsung has both in its arsenal, relying on a partnership with Marvell Technology, another chip developer, for the purpose-built stuff. What would have been more interesting than a demonstration solely of virtual RAN is a comparison of Intel's central processing units (CPUs) with the custom chips designed by Marvell on various criteria – and not just the peak speeds achieved in a lab setting.

Better yet would be a long overdue analysis measuring Intel's virtual RAN processors against a flavor of virtualization called "inline," which splits RAN functions between a general-purpose processor (probably supplied by Intel) and a more customized chip (often described as an accelerator) for the more demanding "Layer 1" category of software. While Samsung, again, appears to have both options available – bringing Marvell in for that Layer 1 offload – its virtual RAN deployments so far look exclusively based on Intel chips. No doubt, it would be uncomfortable about publicly airing the pros and cons of different suppliers.

The job, then, should fall to a major telco, and yet none has had much to say on the matter. Vodafone is known to be examining different chip technologies at its facilities in Malaga. Yago Tenorio, during his tenure as head of network strategy for the UK-headquartered telco, had hinted at the possibility of publicizing results from different virtual RAN trials. Last year, however, he left Vodafone to become Verizon's chief technology officer, and Verizon is heavily dependent on Samsung and Intel for its virtual RAN deployment.

Eclipsed by Nvidia?

In the meantime, the data suggest most operators are still not convinced virtual RAN is worth the effort. Omdia, an analyst company owned by Informa, also Light Reading's parent, put the technology's share of the total market for RAN baseband products at just 10% in 2023 and expects this to double by 2028. It is growing, then, but it remains, and will seemingly continue to remain, far from being the default architectural choice.

Why virtualize? Among the supposed benefits is the ability for RAN developers to exploit the much bigger economies of scale found in the mainstream IT market. "General-purpose technology will eventually have so much investment in it that it will outpace custom silicon," said Sachin Katti, the general manager of Intel's network and edge group, during a previous interview. On the surface, this makes a lot of sense. The whole market for RAN products generated revenues of just $40 billion in 2023. Intel alone made $54.2 billion in sales that same year.

Yet Huawei, Ericsson and Nokia, the big players in RAN technology, have continued to miniaturize and advance their custom chips. Nokia boasts 5-nanometer chips in its latest products and last year lured Derek Urbaniak, a highly regarded semiconductor expert, from Ericsson in a sign it wants to play an even bigger role in chip development.

Ericsson collaborates closely with Intel on virtual RAN, and yet it has repeatedly insisted its application-specific integrated circuits perform better than Intel's CPUs in 5G. This time last year, Michael Begley, Ericsson's head of RAN compute, told Light Reading that "purpose-built hardware will continue to be the most energy-efficient and compact hardware for radio site deployments going forward."

Intel, meanwhile, previously suffered delays when moving to smaller designs and there is gloominess about its prospects. It swung to a $17 billion loss for the quarter ending in September, after reporting a small $300 million profit a year before. Sales fell 6% year-over-year, to $13.3 billion, over this same period.

Unfortunately, for telcos eyeing virtualization, Intel is all they really have. Its dominance of the small market for virtual RAN has not been weakened in the last couple of years, leaving operators with no viable alternatives. This was made apparent in a recent blog post by Ericsson, which listed Intel as the only commercial-grade chip solution for virtual RAN. AMD was at the "active engagement" stage, said Ericsson (this was November). Processors based on the blueprints of Arm, a UK-based chip designer, were not even mentioned.

This could all quickly change if Intel's problems give the industry a big reason to get behind AMD and Arm. But the same economies-of-scale case for virtual RAN is now being made about Nvidia and its graphical processing units (GPUs), which Nvidia boss Jensen Huang seems eager to pitch as a kind of general-purpose AI successor to more humdrum CPUs. If the RAN market is too small, and its developers must ride in the slipstream of a much bigger market, Nvidia and its burgeoning ecosystem may seem a safer bet than Intel. And the GPU maker already has a RAN pitch, including a lineup of Arm-based CPUs to host some of the RAN software.

White water ahead

Semiconductor-related economies of scale, however, should not be the sole benefit of a virtual RAN. "With a lot of the work that's been done around orchestration, you can deploy new software to hundreds of sites in a couple of hours in a way that was not feasible before," said Alok Shah of Samsung Electronics. Architecturally, virtualization should allow an operator to host its RAN on the same cloud-computing infrastructure used for other telco and IT workloads. With a purpose-built RAN, an operator would be using multiple infrastructure platforms.

The trouble is that in markets without much fiber or fronthaul infrastructure there is unlikely to be much centralization of RAN compute. This necessitates the deployment of servers at mast sites, where it is hard to see them being used for anything but the RAN. Even if a company wanted to host other applications at a mobile site, the processing power of Sapphire Rapids, the latest Intel generation, is fully consumed by the functions of the virtual distributed unit (vDU), according to Shah. "I would say the vDU function is kind of swallowing up the whole server," he said.

Indeed, for all the talk of total cost of ownership (TCO) savings, some deployments of Sapphire Rapids have even had to feature two servers at a site to support a full 5G service, according to Paul Miller, the chief technology officer of Wind River, which provides the cloud-computing platform for Samsung's virtual RAN in Verizon's network.

Miller expects that to change with Granite Rapids, the forthcoming successor technology to Sapphire Rapids. "It's going to be a bit of a sea change for the network from a TCO perspective – that you may be able to get things that took two servers previously, like low-band and mid-band 5G, onto a single server," he said.

Samsung's Shah is hopeful Granite Rapids will even free up compute capacity for other types of applications. "We'll have to see how that plays out, but the opportunity is there, I think, in the future, as we get to that next generation of compute." In the absence of many alternative processor platforms, especially for telcos rejecting the inline virtual RAN approach, Intel will be under pressure to make sure the journey for Granite Rapids is less turbulent than it sounds.