Skip to Content
SAP Sponsored Academic Research Conference at CHM highlighted the latest in academia-led innovation

On August 21, the SAP Office of the Chief Scientist hosted the third annual SAP Academic Research Conference in Silicon Valley. At the conference, leaders in academia discussed joint research projects with SAP in areas like multicore, security, sustainability, computational logic, IT productivity, and supply chain risk.

The Academic Research Conference is a key component of the SAP Sponsored Academic Research Program and highlights the important contribution of the Labs toward innovation at SAP.

A word about the Sponsored Academic Research Program; it offers LOBs the unique opportunity to connect to academia and harvest leading edge technologies from top North American and Chinese universities. Anyone at SAP can benefit from academic research and lead a project that shows by proof of concept how we can use innovation from academia for our products.


The conference featured two executive keynotes. Hasso Plattner, co-founder and chairman of the SAP Supervisory Board talked about Enterprise Applications – OLTP and OLAP Share One Database Architecture pdf and Ike Nassi, SAP Chief Scientist talked about Enterprise – The Next Generation pdf


Hasso talked about in memory storage for future enterprise apps. As usual, his talk was mesmerizing. Hasso got great reviews from press and bloggers; for example from Josh Greenbaum of Enterprise Matters Throwing Down The Gauntlet – SAP Get Transactional with In Memory – Watch Out Oracle, IBM and MS – SAP Wants to Change the Game. Josh wrote, “Last week at a conference hosted by SAP to showcase its research work with the academic community, supervisory chair and permanent SAP visionary Hasso Plattner took the concept to its ultimate conclusion: an open attack on the sacred relational database that powers three of SAP’s major rivals/partners and forms the technology core of every SAP installation on the planet”.

Or from Oliver Marks of ZDNET Hasso Plattner Demos Agile Columnar Database Concepts, “Plattner envisages a future world where executives sit in a business ‘cockpit’ combining the human-to-human communication strengths of Cisco’s Telepresence with sophisticated contextual data dialed up on the fly in a much more agile future SAP enterprise infrastructure”.

The discussion between Hasso and Dave Patterson than ensued Hasso’s statement, “we don’t need indices anymore” was without exaggeration epochal; remember, a common agreement at SIGMOD was that indices are one of the main differences between RDBMS and data storage in the cloud.


Ike waltzed on stage to Bob Dylan’s song, The Times They Are A-Changin’… He shined a headlight on the road ahead to the next generation enterprise software and gave the idea how to get there. Write a new layer of SW on top of legacy systems and make sure that he additional data and behavior is built directly into the new layer. The integration capabilities of the layer will allow existing applications to work without change. The new layer will deliver the Enterprise X.0 capabilities that support integration and migration of data to and from legacy systems. Most importantly, the new layer is designed to be timeless supporting changes in data schemas, system infrastructure and applications. You may think this is a tall order. Yes, but it has been done before by MS -from DOS to WNT- and Apple -from PowerPC to Intel platform-.

Further, Ike pointed out that the digital natives will expect to use context for biz apps in the same awesome way they have been using it for web apps before entering the workforce. What about using Hasso’s real time analytics to predict what data the user would like to see next?


Vishal Sikka, CTO of SAP introduced Ike and Hasso. During Hasso’s intro he made some very good points about applications driving the systems requirements and not the other way around; the hammer pushes the nail.


Dave Patterson, professor at UC Berkley, co-inventor of the SPARC chip and RAID technology talked about Cloud Computing and the RAD Lab pdf. Besides giving a terrific and comprehensive definition of cloud computing, Dave presented the work of his team at RAD Lab; machine learning techniques for load management in the cloud. Hasso stated in a direct answer to Dave that he does not believe in higher server utilization; the usual 10 to 20% is the best that can be done; I would add – at least for biz apps. I liked Dave’s broad wing span from hard core CS to the economics of cloud computing; a very unique approach in academia.


Erik Brynjolfsson, professor at MIT Sloan School of Management talked in his presentation When Software is Not Enough: Synergies between Human Resource Practices and Human Capital Management Software pdf about how business practices have to support IT systems. Erik who is Director of the MIT Center for eBusiness has written many books and articles about how IT investment increases the market capitalization for companies. His last book “Wired for Innovation: How Information Technology is Reshaping the Economy” will feature a quote from our CEO Leo Apotheker on the book cover. The book is a short, sharp look on how IT has affected productivity, innovation, competition, pricing, consumer surplus and other aspects of the economy.


Peter Jackson, professor at Cornell University and one of the top researchers on operations research in the world talked about IT Strategies for Increased Rail Employee Satisfaction pdf.Collaborative Data Management pdf.  His students won a price for their project with SAP, Axon and Canadian Rail. This collaboration with Cornell helped SAP to place an ERP contract with Canadian Rail (a light house customer in the North American rail industry).


John Williams, professor at MIT and head of the Auto ID Lab (the Lab that invented RFID) presented the Next Generation Software Systems for Smart Grid and Smart City pdf.  John was named one of the 50 most influential people in computer networking by Network World Magazine. John said that smart meters will generate 1 zeta byte (a 10 with 22 zeros) of data per year. “If one observation can be made about the enterprise of the future, it’s that data generation and analysis will rise dramatically, in ways we are barely capable of understanding, let alone knowledgeable about managing. We no longer collect data simply to have it “just in case” but so we can act upon it, in real time”, wrote Ida Rose Sylvester from Silicon Angle in SAP and The Next Generation Enterprise: Head in the Clouds?

John made the point; the only way to understand how these gazillions of data can be consumed using multicore/multithread computing is to simulate the 1 billion meter reads per day (and per utility company). The simulation enables us to learn how the real time meter data interact with the back end data (billing information, prices, etc) from the ERP systems of the utility companies.


Michael Genesereth, professor at Stanford University and co-founder of CommerceNet and Mergent Systems, an early vendor of technology for integrated catalogs on the World-Wide Web which was sold to Commerce One in 2000 talked about Collaborative Data Management pdf. He is the director of the Logic Group at Stanford and research director of CodeX, the Stanford Center for Computers and Law. He talked about the next generation World Wide Web, the World Data Web. In the World Data Web one will receive answers to queries instead of documents. Collaborative data management systems allow for web queries, not only for flat text search Google style. Concepts and relationships between data in the Web (virtual semantic net) will enable a single data entry principle with a defined data ownership despite the fact that the data pertinent to person e.g. is dispersed over many documents with different ownerships. When sources are independent, they can be updated independently. In the face of interrelationships among sources, individuals performing updates must collaborate (explicitly or implicitly) to ensure correct updates. Collaborative Data Management will replace independent data management.


Blake Johnson, professor at Stanford gave a presentation on It’s an Uncertain World, but Our Planning Capabilities are Deterministic: Making the Transition to Range (Stochastic) Planning pdf. Blake used in this project his know how as former manager of asset risks at Group of Credit Suisse First Boston to evaluate how uncertainty propagates along the supply chain for manufacturing companies. Analyzing many companies like Applied Materials, HP, Cisco, Intel, Dell, Motorola, Genentech and others he found that understanding uncertainties in sourcing and outsourcing relationships offer the largest opportunity for these companies. The pilot company could increase its gross margin by 30% with range planning. They shipped the sure demand and flew only the short lead time demand with high variation/uncertainty. They have saved transportation costs by using shipping over the ocean instead of flying all goods via plane as they did before the study. The savings in transportation cost were much higher than the higher inventory costs due to the longer lead time of ocean shipping.


Gerhard Fischer, Director of the Center for Lifelong Learning and Design (L3D) at the University of Colorado at Boulder and a member of the Computer Human Interaction (CHI) Academy talked about An Analytic Framework for Understanding and Fostering Peer-Support Communities in Using and Evolving Software Products pdf. Gerhard showed his vision of peer support communities using his analysis of SAP Community Network (SCN). He argued that most technical innovations will only be truly effective, innovative, and transformative if equal attention is paid to changes in human behavior. He argued convincingly that the rise in social computing has facilitated a shift from consumer cultures (specialized in producing finished goods to be consumed passively) to cultures of participation (in which all people are provided with the means to participate actively in personally meaningful problems). These developments represent unique and fundamental opportunities and challenges for not only increasing productivity but have the potential to lead to new levels of creativity. By comparing SCN with two open source communities, his results showed that the communities working with SAP products earned high marks for responsiveness and engagement intensity. In addition, his research critically evaluated the reward system which was introduced in SCN a number of years ago.


Jeff Stylos, recent PhD graduate from CMU, presented the work on API Usability for Web Services pdf. This work, led by Professor Brad Myers, discussed studies of developers and APIs on what makes development difficult, along with new tools to help developers. This ispart of the larger Natural Programming Project. He made the point that developers are people, so well-established techniques from Human-Computer Interaction (HCI) can be used to discover important and valid data to drive design. For example, his studies have shown improvements by factors of 2 to 11 times faster by using HCI principles to design APIs. Two new tools have resulted in part from the SAP-sponsored research: Jadeite looks like JavaDoc but adds examples automatically extracted from the web, different font sizes to emphasize the most commonly used items, and crowd-sourced “placeholders” for methods that seem like they should be there but are not. Apatite provides a novel interface for showing items that are commonly associated with other items, and allows developers to explore starting from actions and properties, not just from classes.

To report this post you need to login first.


You must be Logged on to comment or reply to a post.

  1. Former Member Post author

    This conference was extraordinary. I was really interested in the presentations by the various university professors and researchers, and of course by Hasso, Ike, and Vishal. I was also excited to see our own SAP Community Network studied and compared (favorably) against two open-source communities. This is really valuable content on a wide range of topics <– highly recommended to other members of our community, whether you’re a technologist, business person, BPX, or other. Thank you for sharing this so openly with our community members and beyond.

    Mark Yolton

  2. Former Member
    The size of servers will shrink, and capacity to store database will increase substantially. The coding requried to Manipulate Database will be minimized. The Three schema Architecture that is Storage Schema, External Schema, Conceptual Schema will be outdated and only single line interface will be developed. The systems will be more and more end user friendly, The implementation cost will reduce, The Hardware requirement will reduce, The consultants required for maintainence and implementation will reduced. The Distributed Database Concept with Global Schema, Fragmentation Schema, Allocation Schema, Location Schema, Local Mapping techniques, Distribution Transperancy, and the Transmission of Data, Speed, Accuracy, will increase, Data Redundency will reduce, Networking cost will reduce with newer technologies and Most of the Reports, figures, charts, Key figure, Characterstic figure will be available without to much intervention with the system. There will be no need to run updation report, no need to maintain table entries, Faster Implementation with just need of Two servers like Sandbox and Production only will be required because of accuracy and robust programming technique. The futures ERP Systems will not be only Data processing or Manipulating systems they will be Artificially Intelligent systems and will be generations ahead of expert system too. They will be able to accommodate any global business process with tailormade configurations, and the concept of Data Modelling, Process Modelling, Object Modelling will be outdated and the concept of Class Interaction, Hirarchy interaction, Process Interaction, user interaction, and voice programming will be introduced with the help of advance Voice Recongnition softwares, so user will only define exactly what he want to implement and the same interpretation will be done by system and there will be optional closely matching configuration will be delivered out of that, the one suits more to the end user will be accepted. The expandibility, Robustness, Consistency, Accuracy, Audibility, enhancement in concurrency reduction in redundency will be improved significantly in the Near future.

Leave a Reply