Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

To celebrate FabCon Vienna, we are offering 50% off select exams. Ends October 3rd. Request your discount now.

Reply
philnotts
New Member

Does Exporting Data Consume Capacity Units?

Hi, 

 

Does anyone know whether exporting data from Power BI visualisation consumes additional Fabric Capacity Units  to the amount of CU used to render the report visuals? We are trying to understand whether exporting could be a major contributor to high capacity consumption. 

 

 

2 ACCEPTED SOLUTIONS
jaineshp
Memorable Member
Memorable Member

Hey @philnotts,

Yes, exporting data from Power BI visuals does consume additional Fabric Capacity Units because it often triggers new queries to fetch the full dataset, not just what's rendered in the visual. This can contribute noticeably to overall capacity usage, especially with large or frequent exports.

Fixed? ✓ Mark it • Share it • Help others!


Best Regards,
Jainesh Poojara | Power BI Developer

View solution in original post

v-tsaipranay
Community Support
Community Support

Hi @philnotts ,

Thank you for reaching out to the Microsoft Fabric Community.

 

As @jaineshp  mentioned, exporting data from Power BI visuals uses additional Fabric Capacity Units (CUs). This happens because the export process usually runs new backend queries to get the full dataset, not just the summarized data shown in the visual.

These queries need more compute resources like CPU and memory, increasing CU consumption. Also, exporting can bypass visual-level filters or limits, such as Top N, which may lead to larger data exports. Frequent exports or those involving large datasets can have a significant effect on overall capacity usage.

To track this, you can check the Capacity Metrics app or review activity logs for export-related loads. In short, exports do use extra CUs and can impact capacity usage, especially in high-demand situations.

Hope this helps. Please reach out for further assistance.

 

Thank you.

View solution in original post

3 REPLIES 3
v-tsaipranay
Community Support
Community Support

Hi @philnotts ,

 

I wanted to follow up on our previous suggestions. We would like to hear back from you to ensure we can assist you further.

 

Thank you.

v-tsaipranay
Community Support
Community Support

Hi @philnotts ,

Thank you for reaching out to the Microsoft Fabric Community.

 

As @jaineshp  mentioned, exporting data from Power BI visuals uses additional Fabric Capacity Units (CUs). This happens because the export process usually runs new backend queries to get the full dataset, not just the summarized data shown in the visual.

These queries need more compute resources like CPU and memory, increasing CU consumption. Also, exporting can bypass visual-level filters or limits, such as Top N, which may lead to larger data exports. Frequent exports or those involving large datasets can have a significant effect on overall capacity usage.

To track this, you can check the Capacity Metrics app or review activity logs for export-related loads. In short, exports do use extra CUs and can impact capacity usage, especially in high-demand situations.

Hope this helps. Please reach out for further assistance.

 

Thank you.

jaineshp
Memorable Member
Memorable Member

Hey @philnotts,

Yes, exporting data from Power BI visuals does consume additional Fabric Capacity Units because it often triggers new queries to fetch the full dataset, not just what's rendered in the visual. This can contribute noticeably to overall capacity usage, especially with large or frequent exports.

Fixed? ✓ Mark it • Share it • Help others!


Best Regards,
Jainesh Poojara | Power BI Developer

Helpful resources

Announcements
September Power BI Update Carousel

Power BI Monthly Update - September 2025

Check out the September 2025 Power BI update to learn about new features.

August 2025 community update carousel

Fabric Community Update - August 2025

Find out what's new and trending in the Fabric community.

Top Solution Authors
Top Kudoed Authors