Centaur: A Chiplet-based, Hybrid Sparse-Dense Accelerator for Personalized Recommendations
Open accessThe DOI matches its Crossref record; 10 of the 10 references it cites were found in a registry or an index.
Abstract
Personalized recommendations are the backbone machine learning (ML) algorithm that powers several important application domains (e.g., ads, e-commerce, etc) serviced from cloud datacenters. Sparse embedding layers are a crucial building block in designing recommendations yet little attention has been paid in properly accelerating this important ML algorithm. This paper first provides a detailed workload characterization on personalized recommendations and identifies two significant performance limiters: memory-intensive embedding layers and compute-intensive multi-layer perceptron (MLP) layers. We then present Centaur, a chiplet-based hybrid sparse-dense accelerator that addresses both the memory throughput challenges of embedding layers and the compute limitations of MLP layers. We implement and demonstrate our proposal on an Intel HARPv2, a package-integrated CPU+FPGA device, which shows a 1.7-17.2× performance speedup and 1.7-19.5× energy efficiency improvement than conventional approaches.
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
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805 · Cited by 73,173
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 54,325
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×
“As a result, the bandwidth-optimized 3D stacked memory employed in GPUs or ML accelerators such as Google TPUs [21], [34] cannot store the embedding tables locally inside their physical memory, preventing them from being used for deploying recommendations.”
BackgroundDeep Neural Networks for YouTube Recommendations
Covington, P., Adams, J., & Sargin, E. (2016) · Proceedings of the 10th ACM Conference on Recommender Systems · doi:10.1145/2959100.2959190 · Cited by 3,539
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
Eyeriss: a spatial architecture for energy-efficient dataflow for convolutional neural networks
Chen, Y.-H., Emer, J., & Sze, V. (2016) · ACM SIGARCH Computer Architecture News · doi:10.1145/3007787.3001177 · Cited by 1,182 · cited 1×
“The MLP control unit employs an outputstationary dataflow [11] which tiles the input and weight matrices in [32 × 32] sizes (to be compatible with the PE’s”
MethodAlwani, M., Chen, H., Ferdman, M., & Milder, P. (2016) · 2016 49th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO) · doi:10.1109/micro.2016.7783725 · Cited by 662 · cited 2×
“potential for flexible acceleration of the constantly evolving ML applications [4], [20], [44], [45], [48], [59], [64]–[66].”
BackgroundApplied Machine Learning at Facebook: A Datacenter Infrastructure Perspective
Hazelwood, K., Bird, S., Brooks, D., Chintala, S., Diril, U., Dzhulgakov, D., Fawzy, M., Jia, B., Jia, Y., Kalro, A., et al. (2018) · 2018 IEEE International Symposium on High Performance Computer Architecture (HPCA) · doi:10.1109/hpca.2018.00059 · Cited by 590
Can FPGAs Beat GPUs in Accelerating Next-Generation Deep Neural Networks?
Nurvitadhi, E., Venkatesh, G., Sim, J., Marr, D., Huang, R., Ong Gee Hock, J., Liew, Y. T., Srivatsan, K., Moss, D., Subhaschandra, S., et al. (2017) · Proceedings of the 2017 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays · doi:10.1145/3020078.3021740 · Cited by 467
A New Golden Age in Computer Architecture: Empowering the Machine-Learning Revolution
Dean, J., Patterson, D., & Young, C. (2018) · IEEE Micro · doi:10.1109/mm.2018.112130030 · Cited by 168 · cited 4×
“While several hyperscalers [18], [23], [23], [25], [27], [54] hint at the scale of compute and memory required to deploy recommendations, little attention has been paid from the computer systems community to address this important research space (e.”
BackgroundInfluential
Cited by
SpAtten: Efficient Sparse Attention Architecture with Cascade Token and Head Pruning
Wang, H., Zhang, Z., & Han, S. (2021) · 2021 IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca51647.2021.00018 · Cited by 427
Splitwise: Efficient Generative LLM Inference Using Phase Splitting
Patel, P., Choukse, E., Zhang, C., Shah, A., Goiri, Í., Maleki, S., & Bianchini, R. (2024) · 2024 ACM/IEEE 51st Annual International Symposium on Computer Architecture (ISCA) · doi:10.1109/isca59077.2024.00019 · Cited by 298
“Prior work exploits such heterogeneity to selectively schedule requests between CPUs and accelerators [38], [52], colocate models with complementary memory usage [30], and partition compute/memory on heterogeneous hardware resources [45], [48].”
BackgroundLazy Batching: An SLA-aware Batching System for Cloud Machine Learning Inference
Choi, Y., Kim, Y., & Rhu, M. (2021) · 2021 IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca51647.2021.00049 · Cited by 69
Self-citationMAGMA: An Optimization Framework for Mapping Multiple DNNs on Multiple Accelerator Cores
Kao, S.-C., & Krishna, T. (2022) · 2022 IEEE International Symposium on High-Performance Computer Architecture (HPCA) · doi:10.1109/hpca53966.2022.00065 · Cited by 58
“For instance, certain sub-accelerators could be optimized for convolutions [11], [66] to service vision models, some for GEMM [41], [67] to service NLP models, and some for and embeddings to service recommendation models [36].”
Background
Metadata
- Authors
- Hwang, R., Kim, T., Kwon, Y., & Rhu, M.
- Venue
- 2020 ACM/IEEE 47th Annual International Symposium on Computer Architecture (ISCA)
- Publisher
- IEEE
- Published
- 2020-05
- Pages
- 968-981
- OpenAlex
- W3042495273
- Cited by
- 99
- References
- 10
- FWCI
- 11.99
- Type
- conference paper
- Language
- en
- Open access
- green
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
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.