OMNIDB towards portable and efficient query processing on parallel CPU/GPU architectures
Driven by the rapid hardware development of parallel CPU/GPU architectures, we have witnessed emerging relational query processing techniques and implemen- tations on those parallel architectures. However, most of those implementations are not portable across di erent architectures, because...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project |
| Language: | English |
| Published: |
2013
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/55094 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | Driven by the rapid hardware development of parallel CPU/GPU architectures,
we have witnessed emerging relational query processing techniques and implemen-
tations on those parallel architectures. However, most of those implementations
are not portable across di erent architectures, because they are usually devel-
oped from scratch and target at a speci c architecture. This project proposes a
kernel-adapter based design (OmniDB), a portable yet e cient query processor
on parallel CPU/GPU architectures. OmniDB attempts to develop an extensible
query processing kernel (qKernel) based on an abstract model for parallel archi-
tectures, and to leverage an architecture-speci c layer (adapter) to make qKernel
be aware of the target architecture. The goal of OmniDB is to maximize the com-
mon functionality in qKernel so that the development and maintenance e orts for
adapters are minimized across di erent architectures. In this project, we demon-
strate our initial e orts in implementing OmniDB, and present the preliminary
results on the portability and e ciency. |
|---|