Power BI Embedded Capacity

[Anonymous]

Hello,

I am working on the integration of Power BI in an application deployed at the company's customers (ISV). So I saw Power BI Embedded. However, I am wondering about the capacity.

NODE : A1

VIRTUAL CORES : 1

MEMORY : 3 GB RAM

FRONTEND / BACKEN CORES : 0.5 / 0.5

PRISE : ~€620.5955/month

With the following characteristics, how many simultaneous users can there be?

Thanks a lot !  

[nickyvv]

Hi @Anonymous,

sorry to say, but "it depends".., on your datasets, reports, how good you follow the Power BI embedded analytics performance best practices.

You can also have a look at capacity planning in Power BI embedded analytics, but the Power BI Capacity Load Assessment Tool that is mentioned there, only works on an existing capacity, so you'll have to buy a capacity first (it can be an A1 at €0.85/hour) and see what load you can put through it.

I think you should be able to have a good understanding in a few days with that tool, so you'll have to invest a few €€ for that.

[Anonymous]

Hi @nickyvv ,

My basic question is, what is 3GB RAM for? Do they contain the data of the different customers? (whereas customer data is supposed to be isolated with workspaces)?

[nickyvv · Accepted solution]

Hi @Anonymous,

the 3GB RAM is to hold the datasets in memory that are actively being queried.

You should really have a look at the Power BI Capacity Load Assessment Tool I mentioned earlier. That way you can determine if the SKU suits your needs.

[Anonymous]

Hi @nickyvv ,

Thank you a lot for you response. I will look at all the links you provided.

I need one more precision.

With 3GB RAM, if a have :

1) report1 based on dataset1 = 2GB

2) report2 based on dataset2 = 1GB

Can I have simultaneously :

- 1 user looking at report1

- 1 user looking at report2 ?

Is it possible to have simultaneously :

-  more than 1 user looking at report1

-  more than 1 user looking at report2 ? (because they are looking the same report based on the same dataset)

What would be the main idea ? 

Thank you again !