Materialized intermediate results dominate memory consumption in query processing, which is a challenge in the age of Big Data workloads. Modern query processors use pipelines to produce materialized intermediate results. The order by which pipelines are executed influence the lifetime of materialized intermediate results, directly impacting the overall consumed memory. By scheduling the execution of pipelines, we can therefore save memory. Our previous work provided a proof-of-concept for pipeline execution ordering to reduce memory consumption during runtime using a simple join query workload. This paper presents the experimental evaluation of an end-to-end implementation prototype called PEOopt within SAP HANA, which generalizes our previous work. For long running queries our prototype reduces the memory footprint by up to 36% without compromising any execution time performance. Hence, we prove the effectiveness of saving memory in real world database systems like SAP HANA.