Achieving Super-Linear Speedup across Multi-FPGA for Real-Time DNN Inference
Open accessThe DOI matches its Crossref record; 7 of the 7 references it cites were found in a registry or an index.
Abstract
Real-time Deep Neural Network (DNN) inference with low-latency requirement has become increasingly important for numerous applications in both cloud computing (e.g., Apple’s Siri) and edge computing (e.g., Google/Waymo’s driverless car). FPGA-based DNN accelerators have demonstrated both superior flexibility and performance; in addition, for real-time inference with low batch size, FPGA is expected to achieve further performance improvement. However, the performance gain from the single-FPGA design is obstructed by the limited on-chip resource. In this paper, we employ multiple FPGAs to cooperatively run DNNs with the objective of achieving super-linear speed-up against single-FPGA design. In implementing such systems, we found two barriers that hinder us from achieving the design goal: (1) the lack of a clear partition scheme for each DNN layer to fully exploit parallelism, and (2) the insufficient bandwidth between the off-chip memory and the accelerator due to the growing size of DNNs. To tackle these issues, we propose a general framework, “Super-LIP”, which can support different kinds of DNNs. In this paper, we take Convolutional Neural Network (CNN) as a vehicle to illustrate Super-LIP. We first formulate an accurate system-level model to support the exploration of best partition schemes. Then, we develop a novel design methodology to effectively alleviate the heavy loads on memory bandwidth by moving traffic from memory bus to inter-FPGA links. We implement Super-LIP based on ZCU102 FPGA boards. Results demonstrate that Super-LIP with 2 FPGAs can achieve 3.48× speedup, compared to the state-of-the-art single-FPGA design. What is more, as the number of FPGAs scales up, the system latency can be further reduced while maintaining high energy efficiency.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
ImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2577031 · Cited by 73,472
You Only Look Once: Unified, Real-Time Object Detection
Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.91 · Cited by 53,803
In-Datacenter Performance Analysis of a Tensor Processing Unit
Jouppi, N. P., Young, C., Patil, N., Patterson, D., Agrawal, G., Bajwa, R., Bates, S., Bhatia, S., Boden, N., Borchers, A., et al. (2017) · Proceedings of the 44th Annual International Symposium on Computer Architecture · doi:10.1145/3079856.3080246 · Cited by 4,537 · cited 1×
“For instance, Google’s TPU [10] requires the batch size of at least 16 for energy efficiency while FPGAs can extract parallelism from individual execution instance to reduce latency for the low or even no batching.”
MethodSelf-citationRecent Trends in Deep Learning Based Natural Language Processing [Review Article]
Young, T., Hazarika, D., Poria, S., & Cambria, E. (2018) · IEEE Computational Intelligence Magazine · doi:10.1109/mci.2018.2840738 · Cited by 2,921
Optimizing FPGA-based Accelerator Design for Deep Convolutional Neural Networks
Zhang, C., Li, P., Sun, G., Guan, Y., Xiao, B., & Cong, J. (2015) · Proceedings of the 2015 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays · doi:10.1145/2684746.2689060 · Cited by 2,158 · cited 7×
“Figure 2 shows the design space exploration of Layer 5 in AlexNet [1] using the model proposed by [14] which is based on roof-line model.”
BackgroundMethodResultInfluentialScaling for edge inference of deep neural networks
Xu, X., Ding, Y., Hu, S. X., Niemier, M., Cong, J., Hu, Y., & Shi, Y. (2018) · Nature Electronics · doi:10.1038/s41928-018-0059-3 · Cited by 462 · cited 1×
“At the early stage [14, 21, 27–29, 34 ], the Most recently, with the growing demand in time performance, it is a trend to employ a cluster of FPGAs to execute DNNs [5, 7–9, 17, 18, 35, 37].”
Self-citation
Cited by
Device-Circuit-Architecture Co-Exploration for Computing-in-Memory Neural Accelerators
Jiang, W., Lou, Q., Yan, Z., Yang, L., Hu, J., Hu, X. S., & Shi, Y. (2021) · IEEE Transactions on Computers · doi:10.1109/tc.2020.2991575 · Cited by 96
Self-citationAlgorithm-hardware Co-design of Attention Mechanism on FPGA Devices
Zhang, X., Wu, Y., Zhou, P., Tang, X., & Hu, J. (2021) · ACM Transactions on Embedded Computing Systems · doi:10.1145/3477002 · Cited by 52
Self-citationQi, P., Song, Y., Peng, H., Huang, S., Zhuge, Q., & Sha, E. H.-M. (2021) · Proceedings of the 2021 Great Lakes Symposium on VLSI · doi:10.1145/3453688.3461739 · Cited by 37
Self-citation
Metadata
- Authors
- Jiang, W., Sha, E. H.-M., Zhang, X., Yang, L., Zhuge, Q., Shi, Y., & Hu, J.
- Venue
- ACM Transactions on Embedded Computing Systems
- Published
- 2019-10-08
- Volume
- 18
- Issue
- 5s
- Pages
- 1-23
- ISSN
- 1539-9087 (ISSN-L)
- 1558-3465
- OpenAlex
- W2962953210
- Cited by
- 80
- References
- 7
- FWCI
- 13.33
- Type
- journal article
- Language
- en
- Open access
- bronze
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Achieving Super-Linear Speedup across Multi-FPGA for Real-Time DNN Inference. https://app.cochat.ai/papers/10.1145/3358192
- The paper
- Jiang, W., Sha, E. H.-M., Zhang, X., Yang, L., Zhuge, Q., Shi, Y., & Hu, J. (2019). Achieving Super-Linear Speedup across Multi-FPGA for Real-Time DNN Inference. ACM Transactions on Embedded Computing Systems. https://doi.org/10.1145/3358192
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.