Hardware accelerated reconfigurable processor for accelerating database operations and queries
Embodiments of the present invention provide a hardware accelerator that assists a host database system in processing its queries. The hardware accelerator comprises special purpose processing elements that are capable of receiving database query/operation tasks in the form of machine code database instructions, execute them in hardware without software, and return the query/operation result back to the host system. For example, table and column descriptors are embedded in the machine code database instructions. For ease of installation, the hardware accelerators employ a standard interconnect, such as a PCle or HT interconnect. The processing elements implement a novel dataflow design and Inter Macro-Op Communication (IMC) data structures to execute the machine code database instructions. The hardware accelerator may also comprise a relatively large memory to enhance the hardware execution of the query/operation tasks requested. The hardware accelerator utilizes hardware-friendly memory addressing, which allows for arithmetic derivation of a physical address from a global database virtual address simply based on a row identifier. The hardware accelerator minimizes memory reads/writes by keeping most intermediate results flowing through IMCs in pipelined and parallel fashion. Furthermore, the hardware accelerator may employ task pipelining and pre-fetch pipelining to enhance its performance.
1. A database hardware accelerator module configured to be coupled with a host system running a database management system via an interconnect bus, said module comprising:
a set of processing elements configured to receive database machine code instructions and execute them in their respective hardware without software and return results back to the host system upon completion, wherein the processing elements are configured with a dataflow architecture;
a self-routing fabric coupling the processing elements, wherein the processing elements communicate with each other based on Inter Macro-Op Communication (IMC) data structures in a pipelined parallel fashion; and
a memory capable of storing gigabyte portions of the database in compressed, column-store format.
2. The database hardware accelerator module of claim 1 , wherein the IMC is a physical data structure in the processing elements.
3. The database hardware accelerator module of claim 1 , wherein the set of processing elements are configured to support task pipelining.
4. The database hardware accelerator module of claim 1 , wherein the set of processing elements are configured to process data fetched from storage while data is being fetched from storage.
5. The database hardware accelerator module of claim 1 , wherein the processing elements access data in the memory based on a physical address that is arithmetically derived from a global database virtual address space.
6. The database hardware accelerator module of claim 1 , wherein the processing elements access data in the memory based on a physical address that is arithmetically derived from a global database virtual address space and an implicit row identifier of the data in the database.
7. The database hardware accelerator module of claim 1 , further comprising a task processor configured to assign database query tasks from the host database system to at least one of a scanning processing element, an indexing processing element, and an associative processing element.
8. The database hardware accelerator module of claim 1 , wherein the database hardware accelerator module is configured as a memory module package.
9. The database hardware accelerator module of claim 1 , wherein the database hardware accelerator module is configured as a rack-mountable unit.