VMworld Las Vegas 2016 could not have come at a worse time for California-based VMware, just before the September 7 nuptials of its New England parent, EMC, and the Texan Dell.
Many are saying, as usual, that the writing is on the wall for VMware: that is has lost its way, that is it the IBM of the Millennial generation. Watching from afar (the less said about being afar the better, but at least my back is healing) offers a slightly different perspective. Not being dazzled by the bright lights of the conference or being subsumed into the cacophony means that you can more clearly see the chaff from the corn and perhaps discern a direction in what at first seems nothing but white noise.
My colleague Edward wrote a roundup piece in which he succinctly highlights some of the issues VMware is facing. My perspective is that VMware was between a rock and a hard place this year with VMworld. Whilst it is true that it was on the whole left alone to just get on with it as part of the old EMC Federation, there is no real guarantee that this will be the case going forward. Further, its release cycle has managed to become a real mess; nothing significant dropped in terms of a release date for Vegas.
The next version of vSphere (rumored to be 6.5) was not shown at Las Vegas. Perhaps this is due for display in Barcelona, thereby enhancing the status of the European edition of the show. This is a relatively decent shout considering that the Barcelona conference is a day longer, finishing on Thursday rather than the more traditional Wednesday. Perhaps this is so that all the US attendees can watch a proper sports match: the Champions League Manchester City vs. Barcelona match will be held in the city on Wednesday night.
That said, it is true that the US event seemed a little light on substance. At the same time, I really feel that the sub-message regarding management of multiple clouds was on point. To become the One Ring is for VMware the best way of bringing the narrative back to it. We need to hear more about what Kit Colbert’s division is doing regarding cloud-native applications, to hear much more about Photon Controller and container integration, and to see some meat on the bones of its Las Vegas messaging. But it also needs to show that it is still relevant to the rest of the world.
VMware has had a torrid time over the last couple of years, with bungled messages and quality control issues with its patches and key products. (I’m looking at you, NSX.) VMware needs to utilize VMworld as the time to fight back. It fired warning shots across the bows in Las Vegas. Unfortunately, the Silicon Valley analysts all believe that everyone is already drinking unicorn milk with their Yeti steak burgers and living in three-generation application nirvana.
The fact is that VMware can, with tweaking, remain relevant. VSAN is moving in the right direction. ESX needs to be rationalized and made cheaper in order to prevent large-scale Hyper-V migration with the release of Hyper-V Server 2016. NSX needs to be multihypervisor aware by default (hint: it was when Nicira was independent). It needs to build on its relationships with networking vendors. Perhaps Dell Technologies could go out and buy a networking company—Arista comes to mind.
It is true that we are approaching an extinction event in our industry, much like that of the mainframe era, for raw infrastructure. However, the fact remains that there are large tracts of IT that are still not virtualized, using legacy applications that are just too core to the business to be sidelined. That is quite an extreme example, but Windows NT 4 and Windows 2000 abound and will do for a significant time. Companies are not going to migrate their ERP and CRM applications overnight.
Dell Technologies needs to relieve VMware of the quarterly reporting cycle by moving it back, over the fullness of time, to a fully private stock basis. This will hopefully release more money for focused R&D.
To sum up: at Barcelona, VMware needs to double down on the multicloud message—it is the correct one—but also must avoid alienating its current customer base. It should get Photon Controller out the door at a sensible price, stabilize its quality control, and deliver value regularly and consistently. It can start its migration to becoming a post-ESX company, in which the value is delivered from its management layers and not its infrastructure. The Age of the Application is nigh, but there is a very long journey until it is dominant—never mind what those hipster SV analysts say.