I love when folks make these claims, comedy at it's best. Let's tackle these shall we.
1. We chose not to "integrate" like VMware in kernel module for the very simple reason of choice and no lock-in. We are the only HCI Platform to allow all 4 major hypervisors, plus Docker, plus Block Services to Bare Metal plus File Services etc. You get the point. Choice is king. No lock-in is king. Gone are the days of the incumbants taking control of your lives. Now you have freedom. BTW, enjoy updating the Hypervisor to update the storage stack then possibly converting your filesystem with live data to take advantage of the next group of features that we've already had. It's fun being the leader.
2. We have Pulse for proactive support, Prism includes rich Analytics and Trending Analysis out of the box along with I/O metrics..etc. We do not look at just storage, btw, like Nimble we are looking and analyzing on the entire platform, Host, VM, Storage, Disk, etc. BTW, we also have machine learning and our own X-Fit algorithm that's patented for Insight into workloads and Capacity MGMT and soon a ton of other features.
3. Why do I need hardware dedupe? There is no advantage here when Intel XEON CPU architecture can accelerate these workloads very easily which we do with each CVM, GLOBALLY across the Cluster in true Distributed fashion. Let's talk Data Reduction shall we? Inline and Post Process Compression using two different algorithms, inline and post process adaptive Dedupe, and our own ECX-Algorithm which we were first to the table with outside of Goolge. We also don't undersize and overpromise like the vendors mentioned. Every sizing exercise I do is with the customer so they can see our Sizing tool and methodology as a true advantage. I've won deals against the others mentioned by the very fact that we right size and we work together with the customer to make sure that happens. I don't just go out and say here's your BoM with 8TB's oh and btw you get 10:1 reduction without me looking at your data so the 80TB's you needed can be handled with this 8TB. I've seen it time and time again, overpromise, underdeliver. As far as i'm concerned I hope the competition continues down that path.
4. Now, let's talk differences ie via Platform:
Nothing touches Prism period for a MGMT interface. I challenge anyway. You can stick with the Plug in architecture with the Web Client..have fun with that. Prism includes one-click simplicity, non-disruptive upgrades to the entire stack AOS, AFS, ACS, Hypervisor, All firmware and we do this globally to multiple clusters and you can choose to stagger or group them. Prism includes rich analytics, log collection and bundling, and SPoG if you choose AHV as the hypervisor. We have an integrated SSP and will be showing more of this new INTEGRATED (NOT BOLT ON) features and functionality at .NEXT in a couple of weeks. There is a lot more which I think will show everyone where we are heading and it will be exciting to be there to be a part of it.
Security. Security is not an afterthought with our Platform. It's fully integrated and we have our own STIG where our DoD customers (and we have many) have provided insight into this. We ship the box fully compliant to the STIG which is based on DoD standards and we have an automation framework that unsecure changes to settings etc by reverting those changes and reporting back to you. When you factor in AHV, that security is extended to our Hypervisor and included.
AHV does not equal KVM period. At the Core is still QEMU and Libvirt, however we have many advancements in MGMT and Storage Awareness that KVM does not have, DRS, etc.
This is why it's a platform. And I only addressed a small subset. I suggest you take a gander at the Nutanix Bible. nutanixbible.com to get a detailed view of our architecture and why it's truly Enterprise Grade and why we call it the Enterprise Cloud Platform.
BTW, we love HPE's marketing attacks against us. Proves that were the ones to beat..i'm proud of that. Funny though, Josh Odgers is addressing all those as well as our customers on the social networks of FUD etc. In case you want to see resiliency, here's a demo: