The Challenges of Power Metrics

IBM announces a 24 Core Power 10 processor testing Oracles per Socket license model.

"Oracle Database SE2 is licensed on servers with a maximum of 2 sockets. The core counts can increase over time without impacting the license obligation. With Oracle Database SE2, customer license costs remain the same regardless of the number of cores in the socket."

In a Statement of General Direction IBM states its intention to announce a high-density 24-core processor for the IBM Power S1014 system (MTM 9105-41B) to address application environments utilizing an Oracle Database with the Standard Edition 2 (SE2) licensing model. It intends to combine a robust compute throughput with the superior reliability and availability features of the IBM Power platform while complying with Oracle Database SE2 licensing guidelines.

Now that has to present a dilemma for Oracle ... it has clearly positioned SE2 licensing as not affected by the number of cores per socket, so how might it view and react such a direct challenge to such advances in processing power and (more aptly) lost revenue opportunities?

This must bring some angst to the halls of Big Red - what to do? The obvious reaction would be to cap the number of cores, however that's a full u-turn on the defining characteristic - and 'promise' - of their licensing terms - "With Oracle Database SE2, customer license costs remain the same regardless of the number of cores in the socket." 

While there would be nothing legally preventing Oracle from refining these rules it would still very much sour and sully the opinions of their customers, and no doubt cause skepticism longer term. And ultimately, it brings to question consistency - might it be more beneficial to everyone to have uniform metrics across vendors? Ultimately it comes down to processing power, so would a universal metric like mainframe MSUs or MIPs be more appropriate?

... we think so.