A few years ago, I told HP’s product manager for public cloud that I thought all public cloud providers would run out of money and get out of the business. I was mostly being controversial to spark conversation. But HP has recently ceased selling its Helion public cloud. While this did prove me right in that case, I’m not sure every public cloud provider will remain unprofitable until it dies. I do think it is very hard to compete with AWS for a commodity public cloud. On the other hand, there are ways to build a public cloud that address clients who will not use AWS.
Let us be clear on what HP has stopped doing. Only the public cloud part of HP’s Helion is gone. This is the bit that competes with AWS and every other place you can rent a VM on the Internet. The public Helion lives at hpcloud.com, and existing customers have until January 31 to migrate off before it is shut down. The relatively short notice period of three months is a concern. Migrating a large environment to a new platform is not a trivial task. It is very interesting that the SLA Helion provided required only two months, but HP chose to provide customers with another month. Public cloud is only one of the ways in which Helion is deployed. Still available are hosted Helion, on-premises Helion, and the telco-grade Helion distribution.
Why did Helion burn out so fast? I think that there are a number of factors at play here. One is that building enterprise data centers is very expensive. The capital cost is huge, particularly to get to the scale that is required to make cloud economics really work. To get to the scale at which you can compete with AWS requires spending a lot of money. Google and Microsoft both appear to be intent on spending that money, but it seems that HP is not. Some of this reluctance may be due to the uncertainty caused by splitting HP into two companies: one for enterprise products and the other for consumer products.
I think a bigger factor is that customers want more than IaaS from public cloud. Public cloud works best when applications are built for it, not when legacy applications are shoehorned into the cloud. These cloud-native applications are usually designed to consume services from the cloud platform. Functions like databases, message queues, and load balancers should be available as services. These services deliver value to developers, who then simply need to code for business logic rather than infrastructure. The AWS training course I attended was all about how to stitch together these services with a minimal amount of code to deliver an application. This is how cloud makes developers productive. Simply delivering VMs with an operating system helps the operations team, but not the developers.
As far as I saw, Helion did not offer these managed services. Helion delivered VMs for customers, who could then build services inside the VMs. This might have succeeded if Helion had allowed easy migration of VMs from on-premises infrastructure without a massive reconfiguration process. It doesn’t appear that this was an option, so migration to the Helion public cloud wasn’t an easy choice. VMware’s vCloud Air uses the same vSphere hypervisor as most on-premises virtual environments, which makes moving to VMware’s cloud easier.
There are customers for public cloud who cannot or will not use AWS but might use a different public cloud. One example is industries that are highly regulated, like financial, pharmaceutical, or medical data industries. General-purpose public cloud is unlikely to suit these customers. A purpose-built public cloud that is compliant with their regulations could be a viable business, particularly since these businesses are often well funded, and thus are able to pay more per VM for a service that delivers what they require. There are also still some legislative constraints on data locations, particularly in Europe, which may drive a need for niche cloud providers. It will be a sign of maturity in public cloud when we see competition for better services rather than simply a race to the bottom to deliver the cheapest VM possible.
I think HP has decided that competing with AWS, Google, and Microsoft is not a path to profitability. I agree with that, and I expect to see more small cloud providers get out of the business. I do hope we see cloud providers that choose to deliver value that AWS does not deliver.