Antonio Neri, CEO of HPE, declared at its Discover event last week that HPE is transforming into a consumption-driven company that will deliver “Everything as a Service” within three years. In addition, Neri put forward the larger concept of “cloudless” computing. Are these announcements a tactical response to the recent wave of public cloud adoption by enterprises, or are they something more strategic?
“Everything as a Service” is Part of a Larger Cloudless Computing Strategy
“Everything as a Service” is, in fact, part of a larger “cloudless” computing strategy that Neri put forth. Cloudless. Do we really need to add yet another term to our technology dictionaries? Yes, we probably do.
“Cloudless” is intentionally jarring, just like the term “serverless”. And just as “serverless” applications actually rely on servers, so also “cloudless” computing will rely on public clouds. The point is not that cloud goes away, but that it will no longer be consumed as a set of walled gardens requiring individual management by enterprises and applications.
Enterprises are indeed migrating to the cloud, massively. Attractions of the cloud include flexibility, scalability of performance and capacity, access to innovation, and its pay-per-use operating cost model. But managing and optimizing the hybrid and multi-cloud estate is challenging on multiple fronts including security, compliance and cost.
Cloudless computing is more than a management layer on top of today’s multi-cloud environment. The cloudless future HPE envisions is one where the walls between the clouds are gone; replaced by a service mesh that will provide an entirely new form of consuming and paying for resources in a truly open marketplace.
Insecure Infrastructure is a Barrier to a Cloudless Future
Insecure infrastructure is a huge issue. We recently learned that more than a dozen of the largest global telecom firms were compromised for as much as seven years without knowing it. This was more than a successful spearfishing expedition. Bad actors compromised the infrastructure at a deeper level. In light of such revelations, how can we safely move toward a cloudless future?
Foundations of a Cloudless Future
Trust based on zero trust. The trust fabric is really about confidence. Confidence that infrastructure is secure. HPE has long participated in the Trusted Computing Group (TCG), developing open standards for hardware-based root of trust technology and the creation of interoperable trusted computing platforms. At HPE they call the result “silicon root of trust” technology. This technology is incorporated into HPE ProLiant Gen10 servers.
Memory-driven computing. Memory-driven computing will be important to cloudless computing because it is necessary for real-time supply chain, customer and financial status integration.
Instrumented infrastructure. Providers of services in the mesh must have an instrumented infrastructure. Providers will use the machine data in multiple ways; including analytics, automation, and billing. After all, you have to see it in order to measure it, manage it and bill for it.
Infrastructure providers have created multiple ways to instrument their systems. Lenovo TruScale measures and bills based on power consumption. In HPE’s case, it uses embedded instrumentation and the resulting machine data for predictive analytics (HPE InfoSight), billing (HPE GreenLake) and cost optimization (HPE Consumption Analytics Portal).
Cloudless Computing Coming Next Year
HPE is well positioned to deliver on the “everything as a service” commitment. It has secure hardware. It has memory-driven composable infrastructure. It has an instrumented infrastructure across the entire enterprise stack. It has InfoSight analytics. It has consumption analytics. If has its Pointnext services group.
However, achieving the larger vision of a cloudless future will involve tearing down some walls with participation from a wide range of participants. Neri acknowledged the challenges, yet promised that HPE will deliver cloudless computing just one year from now. Stay tuned.