Polaris vs Hypermodel

Hello!

 

Exciting new product releases coming from Anaplan announced at Anaplan Live! 2021. One of things a lot of us are curious about and want to know more about is Polaris. What is the difference between Polaris vs a Hypermodel?

 

- Hypermodel gives users the ability to have up to 700GB's of space in a workspace 

- Polaris is an enhanced calculation engine that can handle up to 10 quintillion cells

 

How does Polaris calculation engine impact workspace space? does your model need to be a Hypermodel in order to support Polaris?

Tagged:

Answers

  • Hi @dkangrga!

     

    As far as I know, there are no technical reasons for not connecting Hypermodels to the new engine. However, the engine has not yet been released and is in development.


    I am sure that Anaplan will first test this bundle and give it to clients if he is sure of the quality of the engine's work on such incomprehensible volumes (w HyperModel).

  • @david.harding 

     

    Thank you so much for that info. So that means there won't be a 2.1 billion cell limit on a particular line item.

    Also will there be an auto switch from Hypermodel to Polaris as and when it becomes GA or Do clients have to ask and pay for it - since Polaris natively is a sparse engine.  I believe clients would be super happy to adopt Polaris - can't wait to work on it.

     

    Misbah

    Miz Logix

  • @Misbah 

     

    While Polaris doesn't have the 2^31 -1 limitation that the current engine does, there is a limitation which is 2^64 -1.  And since the models for Polaris should be architected differently than HyperModels and Classic models as in you will not need to do the concatenations flatting multiple lists into one, Polaris will require a rebuild of the model.  Lastly, yes, clients will have to pay for it, but I am aware of the costs.

     

    Rob

  • To add slightly to what Rob said: the 2^64 (roughly 18 quintillion) cell limit in Polaris is *per line item*. There is nothing to stop you having multiple 2^64-1 cell line items, and if they each only have one populated cell - that is still only ~24 bytes (plus some overhead) each.

  • As I understand from all the white-papers and material, you cannot just switch the model from the classic to Polaris as it needs rebuild because it does not support some formula's such as FInditems etc.

  • I see that you stated "standard Polaris workspaces (up to 130GB)", does this mean we need to have workspaces that Polaris specific and import our models into these workspaces or can we keep our legacy workspaces?

     

  • @coraparks 

     

    The switch for defining a Polaris vs. Classic is at the workspace level, not the model level.  So, you would need to create an additional workspace for Polaris models.

     

    Rob