This time we’re going bigger than ever. Fabric, Power BI, SQL, AI and more. We're covering it all. You won't want to miss it.
Learn moreDid you hear? There's a new SQL AI Developer certification (DP-800). Start preparing now and be one of the first to get certified. Register now
Earlier this year, we announced Power BI as a superset of Azure Analysis Services. It presents our investment and vision for the Power BI ecosystem to meet the needs of customers requiring BI solutions that scale to petabytes, are secure, easy to manage, and accessible to all users across their entire enterprise. At the core of this vision is that models support blazing fast performance for interactive analysis and business-critical decisions even over vast data volumes. We’re excited to announce on-demand load capabilities to further ensure highly performant and reliable large datasets. With on-demand load, Large dataset mode enabled models are hydrated into memory faster, which can provide significantly improved report performance.
Based on a sample of datasets our team used for internal benchmarking, we found on-demand load can significantly improve the downstream user experience — observing up to a 38% reduction in report load times!
Announcing_on-demand_loading_capabilities_for_large_models_in_Power_BI
This new functionality is most significant with larger models. For small datasets, load times typically aren’t a problem. However, with large models exceeding 5GB, the large data volumes can begin to affect performance. With on-demand load in Power BI Premium, the query performance of your largest datasets, even when previously evicted due to user inactivity can be blazing fast!
To learn more about the new column segment states, see Discover Storage Table Column Segments.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.