Hyper-Convergence it aint a passing fad.

Hyper-Converged Infrastructure (HCI) is all about getting “out of the infrastructure business”, and focus on the business – just like Converged Infrastructure (CI).   But, HCI is different from CI in several important ways:

1. You can start smaller.   Much smaller.
2. You scale in smaller increments – which “de-risks” sizing small, and growing as workloads get added (and thus adding only what you need).   This is fundamentally different economic curve.  
3. It’s operationally simpler.  Much simpler.

In our experience, a conservative observations is that HCI generally has about a lower capital cost than CI, but more importantly a big improvement (2x lower) operational/administrative operational cost.   This is particularly impactful as there is no “migration event”, you can add/update/remove nodes and capacity easily and non-disruptively.

Like anything that is growing fast, there’s no doubt a bit of a hype-cycle around HCI – and at my last count 28 startups (I wager we will see HCI startup armageddon  in 2017 – it’s just not sustainable… any takers?)

So let’s deflate the hype-cycle so we can get moving in a more practical way: 

• No, HCI will not solve world hunger.
• No, HCI will not bring peace in the middle east.  
• No, HCI is not ALWAYS lower capital cost than CI (at some compute/memory/IO/storage scaling points, CI is better)
• No, HCI cannot support ALL workloads… but it can support most.

Today, the CI market is about 4x larger than the HCI market – but the HCI market is growing faster.   My own data (as one of the leaders in these segments) matches what IDC is showing on this front.

Now, I don’t believe that any one HCI stack (or SDS stack – as I discussed here) will dominate.   There will be no “one HCI to rule them all”.   Why?  Simple.

You cannot simultaneous design an HCI that is:

• … a perfect fit for a customer with 100 VMs AND…
• … scales to 100,000 VMs AND…
• … is super integrated with VMware AND…
• … also super-integrated with Openstack, Hyper-V AND…
• … optimized for the world of cloud native apps and containers.

You can of course hit those (and other) use cases – but you end up with a “flying boat” – something that is less than perfect in any of the dimensions.

That’s why I deeply believe that just like in other markets, an HCI portfolio will be the winning approach (in fact, a portfolio that also includes CI also).

You can summarize our HCI portfolio in a few sentences:

• VxRail = the right choice for customers uniquely focused and standardized on vSphere that want to start small and scale pretty darn big.
• VxRack SDDC = the right choice for customers uniquely focused and standardized on vSphere that want to start small-ish and scales REALLY big.
• VxRack FLEX = the right choice for customer who want a heterogenous rack-scale HCI – something that works well with vSphere, with Hyper-V, with Linux/KVM, containers you name it.    That heterogenous nature means it doesn’t bolt into any one stack – but is horizontally applicable.  It starts small-ish and scales ENORMOUSLY.
• VxRack Neutrino = the right choice for a customer who is uniquely focused on the cloud-native world.

There’s a lot more to each of these – but more than even for a single long-winded Virtual Geek post.   There’s a webcast series through August and September (recordings will also be made available) that covers these in more details.   Click on the pic to go to the registration page

It’s a good time to get educated about HCI – I guarantee that regardless of your industry vertical, size, workloads – HCI will play a growing role in your IT environment, along with Public Cloud and SaaS.    We’ll also be doing a TON at VMworld – with HCI and CI… Hope to see you at VMworld, and on the webcast series!!!