Hippogriff: Efficiently moving data in heterogeneous computing systems

Liu, Yang; Tseng, Hung-Wei; Gahagan, Mark; Li, Jing; Jin, Yanqin; Swanson, Steven

doi:10.1109/iccd.2016.7753307

Cited by 9 publications

(2 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With GPUs reducing the computation time but the increasing volume of datasets, the data movement overhead becomes more significant to the degree that DB engines must be aware [11,70]. Several GPUDB systems incorporate GPU RDMA techniques [4,46,53,63,78,92] to directly access data on the storage devices [15,52,93] or efficiently exchange data among multiple GPUs [55], bypassing the host system's main memory. This paper is orthogonal but will receive significant benefit from this line of research projects.…”

Section: Related Workmentioning

confidence: 99%

TCUDB: Accelerating Database with Tensor Processors

Hu¹,

Li²,

Tseng³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

The emergence of novel hardware accelerators has powered the tremendous growth of machine learning in recent years. These accelerators deliver incomparable performance gains in processing high-volume matrix operators, particularly matrix multiplication, a core component of neural network training and inference. In this work, we explored opportunities of accelerating database systems using NVIDIA's Tensor Core Units (TCUs). We present TCUDB, a TCU-accelerated query engine processing a set of query operators including natural joins and group-by aggregates as matrix operators within TCUs. Matrix multiplication was considered inefficient in the past; however, this strategy has remained largely unexplored in conventional GPU-based databases, which primarily rely on vector or scalar processing. We demonstrate the significant performance gain of TCUDB in a range of real-world applications including entity matching, graph query processing, and matrix-based data analytics. TCUDB achieves up to 288× speedup compared to a baseline GPU-based query engine.

show abstract

Section: Related Workmentioning

confidence: 99%

TCUDB: Accelerating Database with Tensor Processors

Hu¹,

Li²,

Tseng³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…With the growth of dataset sizes, data movement becomes an increasingly signicant overhead when executing applications [11,21,43]. With improvements of non-volatile memory technologies enabling rich bandwidth inside data storage devices, recent research projects, including Summarizer, started to revisit the idea of pushing computation near storage.…”

Section: Related Workmentioning

confidence: 99%

Summarizer

Koo

Matam

et al. 2017

Proceedings of the 50th Annual IEEE/ACM International Symposium on Microarchitecture

Self Cite

View full text Add to dashboard Cite

Modern data center solid state drives (SSDs) integrate multiple general-purpose embedded cores to manage ash translation layer, garbage collection, wear-leveling, and etc., to improve the performance and the reliability of SSDs. As the performance of these cores steadily improves there are opportunities to repurpose these cores to perform application driven computations on stored data, with the aim of reducing the communication between the host processor and the SSD. Reducing host-SSD bandwidth demand cuts down the I/O time which is a bottleneck for many applications operating on large data sets. However, the embedded core performance is still signicantly lower than the host processor, as generally wimpy embedded cores are used within SSD for cost eective reasons. So there is a trade-o between the computation overhead associated with near SSD processing and the reduction in communication overhead to the host system. In this work, we design a set of application programming interfaces (APIs) that can be used by the host application to ooad a data intensive task to the SSD processor. We describe how these APIs can be implemented by simple modications to the existing Non-Volatile Memory Express (NVMe) command interface between the host and the SSD processor. We then quantify the computation versus communication tradeos for near storage computing using applications from two important domains, namely data analytics and data integration. Using a fully functional SSD evaluation platform we perform design space exploration of our proposed approach by varying the bandwidth and computation capabilities of the SSD processor. We evaluate static and dynamic approaches for dividing the work between the host and SSD processor, and ⇤ Gunjae and Kiran contributed equally to the paper.

show abstract