0

In a Google spreadsheet that performs a lot of intense calculation, does opening the sheet on multiple computers provide more CPU resources to spread out the computation load?

The sheet contains big lists (largely generated from data inputs) and mostly these functions: arrayformula, filter, if, sum, and consequential sum where 1 change at the beginning of the chain and propagate chain reactions and cause many other recalculations. But no RAND or NOW, and no importrange however there's multiple sheets in the same file referencing each other.

If data is uploaded from my PC and the status shows "All chances saved in Drive", can I close the tab and expect the other client PCs to perform the calculations, then return later to see the completed results? Or will my PC recalculate everything upon reopening (even without any new changes) the tab?

Does the account signed in on each computer matter? Same account or multiple different ones work better?

PS: One interesting observation I have is that while making "intense" changes on either PC, it can cause spikes in CPU load in the other PCs and sometimes cause them to become unresponsive or hang!?

6
  • What do you mean by "intense calculation"? Is this made only by built-in functions that are able to run offline like SUM, IF or this include functions that only run online like IMPORTRANGE and custom functions?
    – Rubén
    Jan 23, 2019 at 13:55
  • Regarding account signe in each computer matter, yes it matters, but regarding Google algorithms to prevent abuse / account security. Anyway, I don't think that open the same spreadsheet on multiple computers will speedup spreadsheet recalculation, actually I think that it could cause slowness because calculation made by each client should be checked/synced.
    – Rubén
    Jan 23, 2019 at 13:57
  • Oh! I was forgetting to ask if your spreadsheet use volatile/undeterministic functions like RAND, NOW
    – Rubén
    Jan 23, 2019 at 14:02
  • The sheet contains big lists and mostly these functions: arrayformula, filter, if, sum, and consequential sum where 1 change at the beginning of the chain and propagate chain reactions and cause many other recalculations. But no RAND or NOW, and no importrange however there's multiple sheets in the same file referencing each other. Jan 23, 2019 at 15:35
  • Please add those details to the question, otherwise it's too broad.
    – Rubén
    Jan 23, 2019 at 16:10

2 Answers 2

1

There is no way to distribute computation load of a single spreadsheet between several end-user computers. Many of the function calculation are made on client-side each time that the spreadsheet is opened and recalculated, meaning the calculations are made on each computer when the spreadsheet is openened/recalculated. Most of the functions are recalculated each time a cell is edited, the exceptions are built-in functions like IMPORTRANGE and custom functions.

Some workarounds are

Remark

  • Recalculation occur every single time the spreadsheet is opened, meaning when the spreadsheet is opened on different computers even on different web browser tabs it will be recalculated on each computer/tab.
2
  • Thank you for the suggestion. Do you have keywords to search for when finding the right GAS? What are the chances the duplicated data across multiple sheets actually cause more load? Also I've added more details to the question. You might want to have some comments on that. Jan 25, 2019 at 8:12
  • @AllenSierra Edited my answer
    – Rubén
    Jan 25, 2019 at 13:48
0

No. There is no such thing as "hive computation" in Google Sheets. Even if you have a personal array of computers, it does not matter, because computation is done per each end unit. What does matter is the computation power of each single end unit. In some cases, if that power is poor then it can actually slow down the whole process of computation.

In other words, if you have two rigs accessing a spreadsheet from same or different accounts, your sheet will not be "faster" compared to a scenario where you use 1 PC.

1
  • What a shame! You'd thing someone like google would have implemented "hive computation" on a common cloud product! Jan 23, 2019 at 15:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.