I enjoy a good comparison of any product. Both hardware and software solutions. That beeing said, to this day I still have not experienced such thing. Every vendor is too busy trying to emphasize their own strengths not caring how narrow they might be and trying to hide features they are not so good at. This makes comparison difficult in most cases. Generally speaking.
If you argue that you are tired of the expenses related to one hypervisor with 8 hosts, and brag about the effectiveness of another that utilizes only 3 hosts to serve the same purpose, the physical hardware compared can not have been the same. That would be like comparing your old 5400rpm laptop harddrive with the newest SSD and arguing that the SSD is so much better. Off course it will be. Intel and many other developers provide the majority of consolidation ratios today regarding CPU at least. That being said, Dell servers can tolerate just as many vm's as Lenovo or HP servers as long as the underlying harware is about the same or supermicro for that matter.
The hypervisor, how it manages physical resources and further how it schedules them to virtual machines can also take some of the credit in performance and consolidation ratios. But today one hypervisor is not twice or three times more effective then another at this job. My believes is that the differences is minor.
So It is complex to calculate the TCO of any solution and one are bound to forget pieces of importance that would influence the TCO substantially either way. TCO is about more then just the cost of a socket license for the hypervisor, much more.
I feel that today is more about closing deals and grasping market shares than having the best interests of the customer in mind. I think one must be better at taking a higher ground and getting a better and more complete view of the environment before making huge investments in infrastructure. When you invest, you often invest for a 3-5 year timeperiod. within that period it is common to adjust the initial investment by adding resources like a node, more ram or disk to existing nodes and so forth. I have been in the market for a while, and I, as well as many of you boys and girls out there, know that you get enormous discounts on a new investment. But when it is time to add more cpu/ram/disk, you rarely see the same discounts. Which again ads to the complexity of calculating TCO. Dependent on what solution you chose and how it has been sized, some have to reinvest earlier then others.
At the moment I am calmly watching the market, trying to get a bearing of where it is taking us. there are many and major players out there today. trying to see where we are going is not easy at the moment. Cloud has also been a buzz word for quite some time. Maybe the era of one or a few big players is past. Maybe the future is just about having an enormous amount of small players to chose from, where you can take your environment and jump form one solution to the other as often as you like at small costs. just like changing your mobile carrier or how you can buy electricity from one supplier one month, and another the next month...
Who knows were we will be tomorrow....
Sorry, That was more then just my two thoughts, but then again....
This reply was meant more as a general reply and not as a reply just for or against Nutanix.
Thanks for a great forum and good discussions!