Merge pull request #122572 from wgbrown/patch-1

Update fabric-capacity.md

Author: Court72

articles/cost-management-billing/reservations/fabric-capacity.md

@@ -120,7 +120,8 @@ The following examples show how the Fabric capacity reservation discount applies
 - **Example 2** - A reservation that's larger than your used capacity. For example, you buy 64 CUs of capacity and you only deploy an F32. In this example, the reservation discount is applied to the F32. For the remaining 32 CUs of unused reservation capacity, if you don't have matching resources for any hour. You lose the reservation quantity for that hour. You can't carry forward unused reserved hours.
 - **Example 3** - A reservation that's smaller than the used capacity. For example, you buy 64 CUs of capacity and you deploy an F128. In this example, your discount is applied to 64 CUs that were used. For the remaining 64 CUs, you pay the pay-as-you-go rate.
 - **Example 4** - A reservation that's the same size as two used capacities that equal the size of the reservation. For example, you buy 64 CUs of capacity and you deploy two F32s. In this example, the discount is applied to all used capacity.
-- **Example 5** - This example explains the relationship between smoothing and reservations. [Smoothing](/power-bi/enterprise/service-premium-smoothing) is enabled for Fabric capacity reservations. Smoothing spreads usage spikes into 24-hour intervals (except for interactive usage such as reports read from Power BI or KQL). Therefore, reservations examine the average CU consumption over a 24-hour interval. You purchase a Fabric capacity reservation of two CUs, and you enable smoothing for Fabric capacity. Assume that your usage spikes to 4 CUs within an hour. You pay the pay-as-you-go rate only if the CU consumption exceeds an average of two CU per hour during the 24-hour interval.
+- **Example 5** - This example explains the relationship between smoothing and reservations. Smoothing is a feature of Fabric and allows spikes in usage to be spread out over time.   Smoothing spreads background process like Spark jobs and sematic model refreshes over a 24-hour interval.  Interactive process like Power BI reports, KQL and SQL queries are spread out of a 10 minute interval.  See the detailed explanation here: [Smoothing](https://support.fabric.microsoft.com/blog/fabric-capacities-everything-you-need-to-know-about-whats-new-and-whats-coming)  
+For example, You purchase a Fabric capacity reservation of two CUs, and assume that your usage spikes to 4 CUs for an hour.  The processes will run and consume 4 CUs, however, the CU usage is spread out of the 24 hours. This feature allows you to purchase for average workload rather than the peak. Please review the link provided to understand the impact of smoothing if you use more CU that available over 24 hours.  
 
 ## Increase the size of a Fabric Capacity reservation
 
@@ -131,4 +132,4 @@ If you want to increase the size of your Fabric capacity reservation, use the ex
 - To learn more about Azure reservations, see the following articles:
   - [What are Azure Reservations?](save-compute-costs-reservations.md)
   - [Manage Azure Reservations](manage-reserved-vm-instance.md)
-  - [Understand Azure Reservations discount](understand-reservation-charges.md)
+  - [Understand Azure Reservations discount](understand-reservation-charges.md)