Today’s market isn’t an easy one for network equipment vendors, especially those focused on service providers. Nokia may have a strategy to navigate this critical space, and if so every vendor needs to consider it. In its current quarter, Nokia beat on all the key financial metrics, and held to its guidance, which is pretty positive. The most significant point is that Nokia’s success comes largely from the service provider space, which most vendors (Juniper, for example) are finding to be a weak sector. What’s going on with Nokia is important to every vendor in the space.
What’s going on, of course, is 5G, and the highest-level truth Nokia’s numbers expose is that 5G is where it’s at, equipment-wise, for operators. I’ve noted many times that 5G’s unique benefit is that there’s a budget for it, and Nokia’s win for the quarter is surely due to its ability to exploit that budget. Yes, Huawei’s problems haven’t hurt them a bit, but if Huawei were the only issue, other equipment vendors should have seen their own pop. It’s only the 5G players that seem to be moving.
In Nokia’s case, the growth of 5G has more-than-offset the decline in investment in 4G technology. It’s contributed strongly (via private wireless, including 4G and 5G) to a significant Nokia sales increase in the enterprise. I think it’s helped Nokia across the board, and that’s surely a signal of 5G’s importance to equipment vendors overall. Nokia is a 5G success story.
That doesn’t mean that Nokia can expect to continue happily riding the 5G wave to success. Nokia and Ericsson, the two 5G network equipment giants other than Huawei, have been jousting with each other and also trying to cope with operator interest in open-model networking for 5G. Nokia has been the most aggressive of all the big telecom vendors in supporting open-model 5G (Huawei has been the least), but open 5G presents a risk if big-name players from outside the telecom space jump in and field something credible. Nokia wants to make sure that doesn’t derail their own success, and it may be working.
There’s no question that an open 5G strategy would be less costly than one based on proprietary technology from a traditional telecom vendor, but operators are a conservative bunch, and they need to trust that a new open approach to 5G would offer them reliable and timely ingredients from which to build their services. To do otherwise would put the operators themselves at competitive risk. While operators have been broadly interested in open-model 5G, they’ve been antsy about three specific things, according to my operator contacts.
Thing One is integration and support. Operators tell me that an open-model 5G strategy would likely involve somewhere between a dozen and two-dozen elements, most of which would be from different, smaller, vendors. This creates a significant risk, from operators’ points of view, that the whole thing will creak, groan, and collapse rather than sing, once assembled. Then, who do you point a finger at?
Thing Two is evolution. The majority of big 5G deals will be made with companies who have already deployed 4G, and who need to come up with a graceful way of evolving their infrastructure from 4G-specific to the 5G model that’s specifically designed to be backward compatible. Operators know the vendors who built their 4G networks, and those vendors aren’t the new 5G open-model players. Is it possible to evolve to an open-model 5G infrastructure and preserve 4G in parallel for as long as needed, which could be very long indeed?
Thing Three is the radio network, meaning explicitly the radio technology. There have been major improvements in the “open 5G radio” story recently with the announcement that the massive-MIMO radio problem is on its way to being solved, but it’s not solved yet. Without a solution available now, and with no promise of just when one can be expected, and with questions on when massive MIMO would be needed, this is a big risk to ask operators to accept. Nokia will be making a major, and they say market-leading, massive MIMO announcement shortly, according to their earnings call.
The obvious import of this is that if you want to make money in the telecom equipment space, you’d better have a good 5G story. Why that’s the case isn’t much of a question either; it’s easier to sell something that’s budgeted than something that isn’t. The thing that is a question is just how vendors could get onto the 5G bandwagon. It’s not as easy as it might seem.
In order to have a useful play in 5G in 2021, you need a play in O-RAN, something that Nokia and rival Ericsson have demonstrated already. That means that you have to field one or more O-RAN elements yourself, create a strategy to host them, or both. It is possible for any major network equipment vendor to create O-RAN elements, but if that task hasn’t been underway for at least nine months, it might take too long to bear fruit this year. In addition, cloud software vendors like VMware have fielded their own strategy for O-RAN, one that will ultimately include both platform software and some specific O-RAN component implementations. The space could get pretty competitive.
More competition can be expected, even from the network router players. Nokia is in an interesting position because they are a credible O-RAN player, and because they also provide network routing gear. Many companies in the router space will find it difficult to eke out a 5G positioning because they have nothing to host it on, and no time to build components for O-RAN. Nokia has a leverageable story in the core, with O-RAN to ease the sales dialog.
Things aren’t as rosy with the other telecom router giants. Cisco alone among the telecom equipment players is a credible provider of hosting platforms, both hardware and software. Their O-RAN strategy is to exploit alliances with others. This strategy has been in place for a year, though, and I’m not seeing a lot of movement or traction. Juniper recently announced licensing O-RAN technology from Netsla/Turk Telecom, but that arrangement is new and isn’t bearing visible fruit yet. Juniper’s Contrail Cloud stuff could be a platform for 5G, of course, but this whole story needs to mature. Thus, neither Cisco nor Juniper are in a position to really exploit 5G, and they may not be in that position in 2021.
Among the telco network routing vendors, startup DriveNets may have an interesting advantage here. Their disaggregated routing model, coupled with their ability to host third-party elements as part of their cluster software, means that they could in theory host 5G O-RAN and/or 5G Core components, and since the routing control plane is already hosted there (along with the user/data plane) the cohabiting of all these elements could generate a more efficient implementation. Juniper cites its own control/data-plane disaggregation credentials when talking about 5G, so they might also have designs on exploiting this angle.
It’s also possible that a software player (Red Hat, VMware, or Wind River, for example) could create an alliance with white-box players and offer a platform solution, or even a model architecture of their own. White boxes running OpenFlow, combined with server-hosted control-plane and O-RAN elements, could create a whole new network model, and would provide an inroad for these vendors with operators whose infrastructure spending is largely focused on 5G.
The reason this is interesting is that we’re also seeing a lot of edge stories, including one from Red Hat, who made a number of “edgy” announcements at their conference. Edge computing, of course, requires hosting as a utility function, but it also requires both a conception of just what edge services are and how they would be architected for deployment. Right now, we don’t have any player touting a good story in either of those spaces, and if the same player grabbed 5G O-RAN positioning power and edge power, they could end up being formidable.
All the network vendors, including Nokia, Cisco, Ericsson, and Juniper, aspire to be an edge player, but this is going to be a major challenge. The edge is a lot more like the cloud than like the network, which means that cloud software vendors may have such a natural advantage here that it will be hard for network vendors to overcome. In fact, I think that only riding O-RAN to an early edge lead could level the playing field for the network vendors.
5G is a major driver of telecom spending, and that’s going to be true for years to come. If edge computing manages to escape the hype phase and deliver actual value propositions, it will be an even greater driver, and 5G O-RAN in particular may be the on-ramp to edge computing, which could make 5G-ready vendors even bigger winners. Nokia needs a bit more strategizing, but it might have a shot.