VDI Calculator v4 launch with support for Hyper-Convergence

While we all ramp up for VMworld 2013 I am excited to announce the release of a new version of my calculator. This new version (4.0) introduces the support for hyper-converged infrastructures.

For years, IT organizations have built data centers organized around three key independent hardware segments – compute, network and storage. However, with the rise of commoditization and the move to automated and software defined architectures,  converged hardware is growing in popularity. These solutions may provide reduced complexity, lower overall costs, and easier upgrade paths than are often found with more traditional solutions.

This version of the calculator introduces an easy way to identify requirements for hyper-converged infrastructures. Please note that vendors may have different architecture models that require fine adjustment to the calculations. Please talk to your vendor.


Screen Shot 2013-08-22 at 10.33.43 AM

To access or download the new calculation click the green button below.



What’s New Release 4.0

  • Support for hyper-converged infrastructure deployments
  • Little fixes to validations


This article was first published by Andre Leibovici (@andreleibovici) at myvirtualcloud.net.


  1. Hi Andre,

    Nice release of an already great tool. I am really amazed with both the quality and quantity of the stuff you are posting!

    One question though: what’s the logic behind the “converged” button?
    Do you really assume there is no difference between front end IOPS and backend IOPS with solutions like vBlock and Flexpod or hyperconverged solutions like Nutanix and Simplivity?

    This could be the case for linked clones + local storage without persistent datastore, but in all other cases there is some kind of write penalty. Some converged solutions are using some kind of RAID, other solutions calculate parity over nodes or just mirrors writes over nodes. In every case there is some write penalty.
    I understand that it’s unfeasible to include all those solutions yourself, maybe you can ad them as sponsored calculation? Assuming no write penalty at all is however unrealistic.

    kind regards,


  2. Hi bvdlatisit,

    The calculator is for hyper-converged infrastructures, such as Nutanix, Simplivity and the upcoming VMware vSAN. For the most part, these solutions have a SSD cache to alleviate the IO going to spindles. However, you might be right in saying that we still need to calculate backend IO for those spindles. I was assuming these scale-out solutions would be using parity over nodes, but they may be also using RAID within each node.

    Would you allow users to selected the RAID option for each individual node?


  3. Thanks for the awesome calculator, makes it real easy!

Leave a Reply