Google Cloud’s acquisition of edge computing software company MobiledgeX for an undisclosed sum has been greeted by some commentators as a failure for the company’s ambition of providing an independent platform for operators that keeps the hyperscalers at arm’s length, coming four years after Deutsche Telekom founded the company early in 2018.
It can certainly be seen as a strategic failure, although it looks like DT had given up on the idea of being independent of hyperscalers some time ago, having conceded total ownership of MobiledgeX in 2020 when the company raised external funding from VMware, Samsung, SK Telekom, Worldwide Technologies and Odin Ventures.
That was presented as a step to dilute ownership among multiple operators to prevent it being seen as a DT-centric platform, although this had only deterred telcos that were direct competitors of DT. That included AT&T and Verizon in the USA, where T-Mobile USA, majority owned by DT, is the other major MNO, and Vodafone, which comes up against DT in the German broadband and mobile markets.
Given that Google Cloud has made MobiledgeX’s software available via open source, it is not necessarily the case that the ambition of providing an independent edge ‘broker’ is dead, since operators are free to deploy the software independently. That does, though, depend to some extent on Google’s plans for the software, which otherwise becomes part of its cloud platform.
The original key step towards that brokerage role came in February 2019 when MobiledgeX was still totally under DT’s control, with the launch of version 3.1 of its Edge-Cloud platform, introducing federation between any multi-access edge computing (MEC) platform that supported emerging standards. DT presented this as a milestone towards a ‘write once, deploy anywhere’ capability for mobile networks.
For MobiledgeX itself, absorption by Google is again not automatically a bad result, because acquisition by a major tech player is in many cases the ultimate destiny of a start-up, with success judged partly on the amount of cash obtained.
For DT it could be seen as a failure if the original ambition was to maintain equal distance from the established hyperscalers, AWS, Google and Microsoft. In the event, DT has become ever more entrenched with Google for edge compute, a relationship likely to deepen further with the MobiledgeX sale, and the German firm seems to lack the ambivalence about being too dependent on a single hyperscaler, expressed by Orange in France, for example.
It was just in February 2022 that DT expanded its partnership with Google in three key areas, one being enhanced mobile messaging services for businesses through DT’s RCS Business Messaging underpinned by Google and Jibe Cloud. The second was that DT’s premium TV service, Magenta TV One, running on Google’s Android TV OS, was now available in Germany and being extended to other European countries.
Most intriguing was the third area of the extended partnership, leading to Germany’s emerging Sovereign Cloud becoming available in the country to cloud customers ahead of schedule. On the face of it there is something oxymoronic about that sovereign cloud given that it was supposed to provide an independent platform for Germany enterprises and organizations insulated from the great hyperscalers. But that ambition quickly dissolved as it would have meant excessive investment and delay playing catch-up, which might never have happened anyway.
What might seem surprising is that Google alone has been entrusted with provision of underlying technology, but this reflects its close relationship with DT. In the event the network has been built jointly by Google Cloud and DT’s systems integrator T-Systems, with Google pledging $1.2bn to the cause over 10 years. This is a departure from the original plan for a European sovereign network called Gaia-X built jointly by Germany and France. Subsequently France, despite or because of also being hitched to the American hyperscalers, has attempted to rein back from such close involvement.
This provides the backdrop to Google’s acquisition of MobiledgeX, whose software will be part of the Sovereign Cloud. Ironically, the initial concern for MobiledgeX founders was that it might be swamped not by hyperscalers, but by DT and then other telcos joining in, which is why it was kept at some arm’s length from DT, as a start-up rooted in Silicon Valley. This was also behind the attempt to find other investors including technology vendors as well as telcos. It turned out that most telcos approached were reluctant to follow up expressions of support with hard cash to acquire a stake, and the company has ended up in the hands of a hyperscaler after having originally hoped to thrive as an independent entity with broader ownership.
Nonetheless, MobiledgeX did win business from around 25 telcos, including Telefónica, BT and Orange, as well as DT itself, for what would once have been described as middleware given the software provided a layer at the edge between the operator’s own domain and the public cloud. This was all part of the distribution of computer resources from centralized data centers to more numerous edge locations, which has become a general trend.
The ambition now might be for the software to become more like the Chrome browser, with Google’s technical support but open source so that customers or third-party developers can tailor it. Google can certainly claim to be a leader in open source as a major contributor to various major platforms, such as the widely used Kubernetes software container system.
Meanwhile, carriers are left in limbo, uncertain over their emerging relationship with the hyperscalers. We hear the likes of Vodafone’s chief digital officer Scott Petty scorning AT&T for its decision to host its 5G network in Microsoft’s Azure cloud, arguing this is outsourcing a core competency. Yet Vodafone itself is closely aligned with Amazon’s AWS, as a partner for MEC services delivered with AWS Wavelength for its business customers in the UK.
Vodafone can argue that it also has partnerships with other cloud vendors, but then so can AT&T, and clearly most tier 1 telcos have outsourced core competencies to some degree.