I am primarily a Business Process Consultant with SAP and along with that I am also involved in the SAP & Intel partnership program as a fellow in our Marketing department. SAP and Intel have been co-innovating for many years on in-memory computing. The result has been an optimized SAP HANA platform with Intel Xeon scalable processors and Intel Optane Persistent Memory, that gives organizations business agility and flexibility with real time analytics and re-imagined performance.
In this blog post, I would explore the impact of Intel Optane as persistent memory on SAP S/4HANA ERP system with business use cases illustrating the benefits of Intel Optane due to its higher memory density, non-volatile data persistence, near DRAM processing speed and yet a significantly lower cost of ownership.
When starting your SAP S/4HANA journey, it’s prudent to focus on your infrastructure to fully optimize your move to SAP S/4HANA and reap maximum benefits out of the innovations. The key to consuming all the innovations that is being delivered and planned, lies in your data management strategy.
The challenge is in keeping the pace of database infrastructure (SAP HANA) in sync with that of the innovations that are being delivered by SAP for SAP S/4HANA ERP. Infrastructure is expensive, especially DRAM, which is crucial for processing of high-volume data at high speed and in-memory as demanded by the new innovations.
Intel Optane presents a new revolution in memory persistency with SAP HANA being the first large-scale data platform to be optimized for such a memory. That means, SAP HANA already knows what kind of data structures would benefit most from persistent memory. It detects the persistent memory which in this case is Intel Optane and adjusts itself by placing the right data structures on persistent memory and leaving the others in DRAM.
Let us also explore some relevant business use cases of including Intel Optane as a persistent memory in the hardware landscape for an organization running on SAP S/4HANA.
Business Continuity with faster re-start times:
DRAM is a low-density memory and the data in DRAM is volatile and moves to the cold tier of the database when there is a planned or unplanned server outage. Upon restart of the server, data moves from the cold tier, (external cold storage, e.g SSD) to the hot tier of the DRAM, thus slowing restart times, especially when the data set is large. When such server outages coincide with critical business functions of financial closings or peak sales seasons, business impact due to system non-availability is significant.
Intel Optane Persistent Memory offers faster restart time as the data stored in the persistent memory is non-volatile and is available for processing transactions as soon as the server is rebooted. For mission critical processes running on SAP S/4HANA, a data reload time of minutes, instead of hours would mean ensuring business continuity after any kind of outage.
Sales & Procurement (Analytics & Reporting):
Business reporting performance is optimized since the data is distributed between DRAM and Intel Optane persistent memory (hot memory layer). Most of the “read” data is now available in persistent memory of Intel Optane rather than the cold layer of disks. This would mean running live reports with a faster response time and with almost real-time values enabling critical business decisions. It would also result in less processing time for mass create, change or delete function e.g. mass creation of Purchase Orders or Sales Orders or mass update to product attributes etc.
More historical data in the persistent memory of Intel Optane would reduce the need for frequent archival, i.e. moving the data to disks (cold layer). This would help in building better predictive models using Machine Learning since significant historical data that needs to be “read” only for predictive modelling would be available in persistent memory.
Human Resource (Analytics & Reporting):
Through embedded analytics, it is now possible to create employee fact sheets with utilization information, availability and time off information. In fact, additional KPI’s like, at the team level can be added, e.g. “team utilization rate” on top of individual KPIs for greater insights. These analytics take advantage of hot data stored in persistent memory of Intel Optane to provide real time reports for effective resource planning and project management.
Manufacturing (Analytics plus AI & ML):
With data persistency, advanced analytics has made it possible to build comprehensive production process model that would provide visibility from production planning right up to the shop floor level. It enables the addition of detailed activity layers that underlines production operations and provides a more granular description of shop floor processes. This in turn, enables live guidance and control over the actual production operation. Proactive monitoring using AI enables planning of independent requirements to help in procure or produce decisions ahead of receiving sales orders when lead times from customers are very short. It also helps to detect forecast errors to cover future demand fluctuations and uncertainty around supplier commitments. With higher memory density in persistent memory of Intel Optane, the AI algorithms can better predict and monitor production planning data across plants and shop floors
Finance with AI & ML:
Using Artificial Intelligence and Machine Learning, predictive data models are available to provide early insights into profit and losses. Predictive accounting can predict journal entries of future goods issue and billing documents by reading date from up to date Sales Orders. Also accelerated financial closing activities are now available for better compliance and risk management. Revenue Recognition issues can be managed real time with the help of scheduled jobs for Sales Order, Projects and Service Documents to accelerate the financial close. These jobs mostly need to perform huge “read” functions across journal entries created out of Sales Order, Project and Service Documents. Intel Optane as persistent memory is best suited to support such large “read” operations.
SAP and Intel have partnered on several initiatives to help you on your SAP S/4HANA MOVE journey. SAP HANA with Intel Optane revolutionizes the capability of SAP S/4HANA as an ERP to fuel the move to an Intelligent Enterprise. This technology gives organizations faster access to large volumes of data with optimized performance, faster re-start times and reduced infrastructure and operational costs. Speak with your SAP or Intel Account Executive to learn more.
For more information on SAP and Intel partnership and further information on Intel Optane, please visit the following sites:
