intel / bdtk Goto Github PK

Hi,

I am familiar with Intel's QPL library and how it interacts with the in-memory analytics accelerator (IAA). So far, I wrote some microbenchmarks and compared the filter performance to internal vectorized filter execution code.

The README and paper for BDTK mention compression acceleration with IAA (but not QPL), and as far as I can tell there is code that references QPL in CMakefiles/wrappers. How do QPL and BDTK interact? Is QPL an official part of BDTK, is BDTK supposed to be a superset of functionality or is QPL a standalone thing that can also be used via BDTK?

I think I understand what Intel's goal is with each project, but the overall strategy of how these efforts fit together is unclear to me. Is there a roadmap or document outlining all of these analytics/query processing/accelerator from Intel?

Thank you,
Jonas