Is Hana Truly Necessary?

It is not unlikely that SAP will reach a dead end with Hana. Further development will occupy a good deal of resources in the coming years that SAP could better implement in the areas of AI, process mining, and the Business Technology Platform.

Hana is an outdated construct and overpriced. Additionally, Hana is based on an outdated computer architecture. When Hana was created in 2011, the first high-performance CPUs were just coming out, main memory was getting cheaper, and disk storage was always slow and error-prone as it was all mechanical. Loading the entire ERP database into the server's main memory was a logical step because of the Von Neumann computer architecture.

John von Neumann (1903 - 1957) was a pioneer in computer architecture. He developed the concept of a central processing unit (CPU) with an attached memory to store both program code and data. This architecture was a huge success. However, the connection between the CPU and memory became a speed-limiting bottleneck. This is now known as the Von Neumann bottleneck.

Professor Hasso Plattner also ran into the von Neumann bottleneck when he tried to compensate for the speed deficit of traditional databases by moving data from slow, hard disks to much faster main memory. Plattner, however, did not recognize the ongoing evolution of hardware technology, nor did he question the prevailing computer architecture. SAP's Hana database is a product of the von Neumann bottleneck. With this in mind, Hana can only be considered hopelessly outdated.

If Professor Plattner and SAP had only wanted to increase speed, to finally live up to the "R" in R/2 and R/3 (real time), a little humility and a few years' wait would have sufficed. But Plattner and SAP wanted to implement real time in 2011. However, what was even more essential to this decision was the political market dimension.

SAP no longer wanted to share customers’ IT budgets with operating system and database providers! Developing its own operating system would be pointless and expensive, but developing its own database to displace Oracle, IBM, and Microsoft was within the realm of possibility. And so Hana was born!

And Professor Hasso Plattner had big plans, as reporter Michael Kroker wrote in WiWo: [...] the US company (Oracle) had earned a lot from SAP's success as a database software supplier for the previous versions of Hana. And so, shortly after the introduction of Hana, Plattner publicly quipped: “I wouldn't want to be in Larry Ellison's shoes right now. After all, a freight train called Hana is barreling toward him.” Ellison countered in typical his fashion: “They must be on drugs”.

In a later YouTube video where Plattner is interviewed, Larry Ellison's "You must be on drugs" allusion is mentioned again. The fact is that Hana did not become a threat to other database vendors, but it did bring quite a lot of revenue directly into SAP's pockets.

Insiders suspect that Hana’s final positioning was the subject of much debate at SAP. Professor Hasso Plattner formulated the goal in an interview with WiWo editor Michael Kroker as follows: "My last tech project that I drove forward at SAP was the high-performance Hana database. I hoped that SAP would succeed in bringing Hana to the top of this area, as it was by far the best database at the time. The market should have rewarded that somehow, but SAP didn't quite manage it.”

Apparently, former SAP CTO Vishal Sikka wanted a public domain version of Hana. Whether Hasso Plattner or the SAP management was against it is not known. The fact is, however, that if SAP had marketed its own database in a similar way to the underlying Linux operating system, it would have missed out on many millions of dollars in licensing revenue.

While Hana may excite only a few computer scientists and drive many CFOs of SAP customers to despair over double database licensing fees, Hana is a resounding commercial success for SAP itself. Now SAP no longer has to share the ERP community with Oracle, IBM, and Microsoft. The entire ERP stack is in SAP's hands. The final vendor lock-in has been crowned with Hana and S/4.

However, it is uncertain whether SAP's calculations will work out in the end, given the ever-increasing pace of IT development. In this case, it would do SAP well to stick to what they know.