# Optimize consumption of high-performance compute
<a name="genperf03"></a>


| GENPERF03: How do you optimize computational resources required for high-performance distributed computation tasks? | 
| --- | 
|   | 

 Foundation models require high processing power to customize and deliver inference at scale. This is true whether you are training a foundation model, customizing it, or using one for inference. Optimize consumption of high-performance compute used for foundation models in order to meet performance requirements. 

**Topics**
+ [GENPERF03-BP01 Use managed solutions for model hosting, customization, and data access where appropriate](genperf03-bp01.md)