Businesses rely on transactional and analytical databases for a single source of truth. Organizations of all sizes need to extract insights from these large databases, and housing the databases in memory can accelerate actionable insights, increasing business competitiveness. But in-memory processing requires massive amounts of DRAM. And while the price of memory continues to decline, single-unit scale-up servers and multi-node scale-out clusters still represent a sizable investment for IT organizations.
With its advanced in-memory technology, the SAP HANA® platform is widely used by organizations that need rapid processing for large volumes of data that is optimized for in-memory workloads. However, the price of acquisition is only part of the cost; SAP HANA deployments can be expensive to operate, particularly when labor costs for the specialists required to manage them are factored in.
To evaluate total system costs, Lenovo commissioned Prowess Consulting to assess scale-up systems in a variety of memory configurations for SAP HANA solutions running on servers powered by the latest-generation Intel® Xeon® Scalable processors from HPE and Lenovo. Our analysis shows that eight-socket Lenovo® ThinkSystem™ SR950 V3 servers running Red Hat® Enterprise Linux® for SAP® Solutions provide a lower total cost of ownership (TCO) over a three-year period, compared to eight-socket HPE® Compute Scale-up Server 3200 servers.1
We found that this lower TCO stems from both lower capital expenditures (CapEx) and especially lower operating expenses (OpEx) for the Lenovo® servers. In addition to raw TCO advantages, the Lenovo solutions for SAP HANA that we evaluated in our study also feature faster times to load databases into memory and perform complex queries than their counterparts from HPE, while providing high levels of reliability. Moreover, at the time of this writing, the ThinkSystem SR950 V3 server, alone among the SAP HANA platform–certified servers we examined, also provides an option for a scale-out configuration, which can be valuable for analytical workloads, such as SAP S/4HANA® software.