Embedded report memory consumption

[j_bourys]

Hey guys!

I have created a basic webpage to distribute embedded Power BI reports on TV for customers, factory workers, administration workers etc.
What solution would you suggest to distribute reports on a TV?

A little back story:
Currently I have embedded small reports (1,5 MB desktop file, 2 tables, 3k rows, 20 measures, 4-10 visuals per page) in a webpage. And there already are problems to open the webpage in browser for TV's that are not so high end, however still a smart TV with streaming apps. Errors like "Not enough memory" are returned.

Most PC easily loads web pages that consumes few hundred MB - it is quite challenge for some TV (Installed RAM ~1GB)

While this is also a hardware issue, because the same webpage with Power BI report works on some TV's and on some does not, have you any thought if the report should be significantly optimized? Is 200MB memory usage a lot for a published embedded report?

[DMiradakis]

I will echo @lbendlin and say that you'll most likely want to try and optimize your Power BI model. Without knowing the specific TV models, it's difficult to say why some TV models can load the report and others can't.

Not sure about the 200 MB question either, unfortunately. I know that my Power BI reports can be significantly large sometimes just depending on the report's use case and whether I'm using a shared dataset, Import Mode, Direct Query, etc.

If you can use a DQ connection to a shared dataset, that might alleviate some of the RAM consumption, but I'm not 100% sure on that.

If you don't end up finding a solution for your problem, I will throw out that I've made a custom Power BI app for digital signage called Displagent. Works cross-platform, too, so if cost is an issue, you can purchase a Raspberry Pi Model 4 that will work very well with it. Has a lot of advanced features beyond just loading a plain report on a screen.

Hope some of that helps!

[lbendlin]

Use DAX Studio to examine the Query Plans for your report visuals, and optimize their memory usage.