F64 memory error

[wojciech]

Hi,

We have just migrated from P1 to F64, and we start getting dataset memory errors for models that used to work just fine in P1 capacity:
"Resource Governing: This operation was canceled because there wasn't enough memory to finish running it. Either reduce the memory footprint of your dataset by doing things such as limiting the amount of imported data, or if using Power BI Premium, increase the memory of the Premium capacity where this dataset is hosted. More details: consumed memory 15806 MB, memory limit 15805 MB, database size before command execution 9794 MB."

Not sure why it's calling out a limit of 15805 while F64 has got a limit of 25GB. Having checked Fabric capacity Metric App we only utilise 10% of overall capacity. 
Would anyone has got any idea what is going on?

Thank you,

WJ

[AntoineW]

Hi @wojciech 

Thanks for raising this — migrating from P SKUs to Microsoft Fabric capacities like F64 often reveals subtle differences in how memory and resources are managed.

To help pinpoint the issue, here are a few key questions to consider:

1) Have you enabled Large Model Format on your dataset?

Go to the workspace setting > License info > fabric capacity > Semantic model storage format and check if:

• Large dataset storage format is enabled

• The dataset is using enhanced metadata format

Without this, your model may be restricted to lower memory thresholds even on F64.

2) Have you reviewed the Capacity Metrics App?

• Are you checking per-dataset memory consumption and peak usage?

• Under Dataset Operations, do you see any throttling or resource contention during refresh times?

• Is your dataset refresh overlapping with other workloads (e.g., AI notebooks, pipelines, other dataset refreshes)?

3) Is your model optimized for Fabric?

• Have you reviewed high-cardinality columns or unused columns in the model?

• Are you using incremental refresh or partitions to limit the memory used during refresh?

• Could you reduce the dataset size with proper data filtering or summary tables?

4) Are you scheduling concurrent heavy workloads on the same capacity?

Fabric enforces isolation and resource quotas more strictly than Power BI Premium P SKUs. Even on F64, a large dataset may only get 15–20 GB of working memory depending on what else is happening on the capacity.

Hope it can help you!
Best regards,
Antoine

[v-menakakota]

Hi @wojciech   ,
Thanks for reaching out to the Microsoft fabric community forum. 

Please go through the solved issue which may help you in resolving the issue:
Solved: Re: Fabric F64 gives an out of memory error on res... - Microsoft Fabric Community

Could you please share error message her.

 

If I misunderstand your needs or you still have problems on it, please feel free to let us know.   

Best Regards, 
Community Support Team  

[v-menakakota · Accepted solution]

Hi @wojciech ,

I hope the information provided above assists you in resolving the issue. If you have any additional questions or concerns, please do not hesitate to contact us. We are here to support you and will be happy to help with any further assistance you may need.

Thank you.