There is a moment in the classic disaster movie story arc when the hero turns to his companions and declares, “Look, I know this is a crazy, bold move, but it’s our only chance.” Huawei has just made such a declaration at its annual global analyst event.
Huawei is expanding from its core telecom equipment market laterally into the larger adjacent markets of enterprise IT and consumer devices. It is also expanding in both directions vertically, relying increasingly on its own silicon supply from subsidiary Hisilicon and attempting to lead its carrier customers into value-added services.
But don’t mistake these directions as simply ambition from a large and growing company. Huawei sees that no existing responses assure that the flood of data traffic can be met with adequate network investment. It has even named its new corporate R&D function “2012 lab,” not only after the current year but also the disaster movie.
Could Huawei 2012 become the next Bell Labs?
Huawei responded to a drop in profits in 2011 by increasingly spending on R&D rather than cutting all expenses. Over the past few years it has continued to hire large numbers of engineers across the board, reaching a total of 62,000 with 11,000 added in 2011 alone. It believes it has matured beyond following or even leading the industry and now seeks to steer the industry direction.
Historically, industry-leading R&D has not been a pure numbers story. Earlier this year at the OFC conference, Huawei made a big push to demonstrate advanced technology both on the show floor and in prestigious post-deadline technical papers. At the time, we speculated that only Huawei is in a position now to take over the industrial research mantle from Bell Labs. We don’t know at the moment how many of the 62,000 R&D staff are part of the corporate 2012 lab. Whatever happens though, the new research lab is already sounding different from the 20th century model where top researchers were free to pursue their own directions. One of the stated purposes of the corporate R&D function, in addition to business unit R&D, is rapid response to changing needs – a sort of hybrid cloud model for engineers if you will.
Silicon and software architecture together are keys to future
The company itself may disagree with the above depiction that it is expanding through vertical integration. Instead it depicts a scenario in which price pressures are compressing the value chain such that the top and bottom layers are the only ones that retain value in the long run. One executive at the event responded to a suggestion that Chinese vendors’ aggressive pricing was “ruining the business model” by rebutting that the telecom business model was already broken and declining equipment prices were the result, not the cause. The only possible response, he suggested, was to use in-house silicon to reduce cost. Another spokesperson said that price pressure for handsets was such that no money could be made on phones, only on the chips.
But margin savings, economies of scale, and Moore’s Law can’t get ahead of a flood which is itself caused by the same enablers. The real potential lies in the ability to make radical innovations unconstrained by somebody else’s silicon or software.
Huawei identifies four specific areas whose chips are critical to its destiny: network processors, photonics integrated devices, passive optical network chips, and wireless chips. Huawei uses its own chips only where it can find differentiation. It continues to use Intel and Broadcom chips, and it was an early adopter of merchant ATCA hardware platforms as well.
Enterprise and consumer businesses are similarly keys to telecom change
The message on the role of Huawei’s enterprise business was mixed. It showed a version of the classic ICT convergence Venn diagram where its current carrier network market lies at the intersection of enterprise and consumer devices. Its product business units are currently organised into these three areas. To investors, the message was that Huawei could leverage its current business for growth particularly into enterprise IT, whose total market size is an order of magnitude larger than its telecom equipment business. Yet market size is not sufficient reason to get into as competitive a market as datacentre servers and switches.
The company acknowledges its new Enterprise business is a work in progress. Teams are trying to learn about specific verticals including public sector, power, finance, and transportation, hoping to find synergies or similarities to the special needs it has learned to address in the telecom vertical. Yet perhaps because it is too early to tell if it will be able to find appropriate special needs, it lists as additional “verticals” large enterprise and SME.
The more credible picture is that the three business units represent the end-to-end networking view where services originate at a consumer device and are provided by a datacentre with the carrier network in between.
In addressing both ends, Huawei would like to help its carrier customers own the datacentres and profitable services rather than just be a pipe provider to over-the-top players. It believes it can make servers and switches that are better optimised for specific customer needs. The server offering makes more sense as an outgrowth from its ATCA servers towards still-customised but more datacentre-centric servers and network equipment.
This picture also puts Huawei in a better light. Instead of graduating into a bigger pond and starting over as just another fish, it becomes the telecom vendor with best end-to-end visibility. Ericsson has mobile but doesn’t play in servers. Cisco has the datacentre network and some devices but not mobile.