Abstract | ||
---|---|---|
Recent advancements in deep learning techniques facilitate intelligent-query support in diverse applications, such as content-based image retrieval and audio texturing. Unlike conventional key-based queries, these intelligent queries lack efficient indexing and require complex compute operations for feature matching. To achieve high-performance intelligent querying against massive datasets, modern computing systems employ GPUs in-conjunction with solid-state drives (SSDs) for fast data access and parallel data processing. However, our characterization with various intelligent-query workloads developed with deep neural networks (DNNs), shows that the storage I/O bandwidth is still the major bottleneck that contributes 56%--90% of the query execution time.
To this end, we present DeepStore, an in-storage accelerator architecture for intelligent queries. It consists of (1) energy-efficient in-storage accelerators designed specifically for supporting DNN-based intelligent queries, under the resource constraints in modern SSD controllers; (2) a similarity-based in-storage query cache to exploit the temporal locality of user queries for further performance improvement; and (3) a lightweight in-storage runtime system working as the query engine, which provides a simple software abstraction to support different types of intelligent queries. DeepStore exploits SSD parallelisms with design space exploration for achieving the maximal energy efficiency for in-storage accelerators. We validate DeepStore design with an SSD simulator, and evaluate it with a variety of vision, text, and audio based intelligent queries. Compared with the state-of-the-art GPU+SSD approach, DeepStore improves the query performance by up to 17.7×, and energy-efficiency by up to 78.6×.
|
Year | DOI | Venue |
---|---|---|
2019 | 10.1145/3352460.3358320 | Proceedings of the 52nd Annual IEEE/ACM International Symposium on Microarchitecture |
Keywords | Field | DocType |
Hardware Accelerators, In-Storage Computing, Information Retrieval, Intelligent Query, Solid-State Drive | Computer architecture,Locality of reference,Computer science,Cache,Search engine indexing,Real-time computing,Artificial intelligence,Solid-state drive,Deep learning,Design space exploration,Data access,Runtime system | Conference |
ISBN | Citations | PageRank |
978-1-4503-6938-1 | 2 | 0.36 |
References | Authors | |
0 | 10 |
Name | Order | Citations | PageRank |
---|---|---|---|
Vikram Sharma Mailthody | 1 | 8 | 1.46 |
Zaid Qureshi | 2 | 5 | 1.06 |
Weixin Liang | 3 | 2 | 1.04 |
Ziyan Feng | 4 | 2 | 0.36 |
Simon Garcia de Gonzalo | 5 | 2 | 0.36 |
Youjie Li | 6 | 13 | 0.96 |
Hubertus Franke | 7 | 1257 | 104.86 |
Xiong Jinjun | 8 | 801 | 86.79 |
Jian Huang | 9 | 2 | 0.36 |
Wen-mei W. Hwu | 10 | 4322 | 511.62 |