GOBO: Quantizing Attention-Based NLP Models for Low Latency and Energy Efficient Inference
Open accessAbstract
Attention-based models have demonstrated remarkable success in various natural language understanding tasks. However, efficient execution remains a challenge for these models which are memory-bound due to their massive number of parameters. We present GOBO, a model quantization technique that compresses the vast majority (typically 99.9%) of the 32-bit floating-point parameters of state-of-the-art BERT models and their variants to 3 bits while maintaining their accuracy. Unlike other quantization methods, GOBO does not require fine-tuning nor retraining to compensate for the quantization error. We present two practical hardware applications of GOBO. In the first GOBO reduces memory storage and traffic and as a result inference latency and energy consumption. This GOBO memory compression mechanism is plug-in compatible with many architectures; we demonstrate it with the TPU, Eyeriss, and an architecture using Tensor Cores-like units. Second, we present a co-designed hardware architecture that also reduces computation. Uniquely, the GOBO architecture maintains most of the weights in 3b even during computation, a property that: (i) makes the processing elements area efficient, allowing us to pack more compute power per unit area, (ii) replaces most multiply-accumulations with additions, and (iii) reduces the off-chip traffic by amplifying on-chip memory capacity.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In every layer, approximately 99.9% of weights follow a Gaussian distribution.
“In every layer, ∼ 99.9% weights follow Gaussian distribution.”
From A Survey of Techniques for Optimizing Transformer Inference (2023) · cites this paper as [126]
GOBO allows memory compression by 10×.
“Their technique allows memory compression by 10 × .”
From A Survey of Techniques for Optimizing Transformer Inference (2023) · cites this paper as [126]
Not storing outliers leads to large accuracy loss, but storing more than 0.1% outliers provides only marginal improvement in accuracy.
“Not storing outliers leads to large accuracy loss, but storing more than 0.1% outliers provides a marginal improvement in accuracy.”
From A Survey of Techniques for Optimizing Transformer Inference (2023) · cites this paper as [126]
Cites
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 48,293 · see citing works
RoBERTa: A Robustly Optimized BERT Pretraining Approach
Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., Levy, O., Lewis, M., Zettlemoyer, L., & Stoyanov, V. (2019) · arXiv · doi:10.48550/arxiv.1907.11692 · Cited by 16,515 · see citing works
DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter
Sanh, V., Debut, L., Chaumond, J., & Wolf, T. (2019) · arXiv · doi:10.48550/arxiv.1910.01108 · Cited by 4,653 · see citing works
SpanBERT: Improving Pre-training by Representing and Predicting Spans
Joshi, M., Chen, D., Liu, Y., Weld, D. S., Zettlemoyer, L., & Levy, O. (2020) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00300 · Cited by 1,827 · see citing works
MobileBERT: a Compact Task-Agnostic BERT for Resource-Limited Devices
Sun, Z., Yu, H., Song, X., Liu, R., Yang, Y., & Zhou, D. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.195 · Cited by 757 · see citing works
SQuAD: 100,000+ Questions for Machine Comprehension of Text
Rajpurkar, P., Zhang, J., Lopyrev, K., & Liang, P. (2016) · arXiv · doi:10.48550/arxiv.1606.05250 · Cited by 678 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
A primer in BERTology: What we know about how BERT works
Rogers, A., Kovaleva, O., & Rumshisky, A. (2020) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00349 · Cited by 1,552 · see citing works
A Review on Edge Large Language Models: Design, Execution, and Applications
Zheng, Y., Chen, Y., Qian, B., Shi, X., Shu, Y., & Chen, J. (2025) · ACM Computing Surveys · doi:10.1145/3719664 · Cited by 169 · see citing works
A Survey of Techniques for Optimizing Transformer Inference
Chitty-Venkata, K. T., Mittal, S., Emani, M., Vishwanath, V., & Somani, A. K. (2023) · Journal of Systems Architecture · doi:10.1016/j.sysarc.2023.102990 · Cited by 154 · see citing works
Utility is in the Eye of the User: A Critique of NLP Leaderboards
Ethayarajh, K., & Jurafsky, D. (2020) · Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2020.emnlp-main.393 · Cited by 147 · see citing works
Dong, P., Zhuang, J., Yang, Z., Ji, S., Li, Y., Xu, D., Huang, H., Hu, J., Jones, A. K., Shi, Y., et al. (2024) · IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems · doi:10.1109/tcad.2024.3443692 · Cited by 35 · see citing works
“If the dataset is not publicly available, users have to use PTQ. QAT requires significant computational resources to support the training of large model sizes (usually over 350M), to which academics usually have limited access.”
MethodA Survey: Collaborative Hardware and Software Design in the Era of Large Language Models
Guo, C., Cheng, F., Du, Z., Kiessling, J., Ku, J., Li, S., Li, Z., Ma, M., Molom-Ochir, T., Morris, B., et al. (2025) · IEEE Circuits and Systems Magazine · doi:10.1109/mcas.2024.3476008 · Cited by 32 · see citing works
Post-Training Quantization for Vision Transformer
Liu, Z., Wang, Y., Han, K., Ma, S., & Gao, W. (2021) · arXiv · doi:10.48550/arxiv.2106.14156 · Cited by 192 · see citing works
- Authors
- Zadeh, A. H., Edo, I., Awad, O. M., & Moshovos, A.
- Affiliations
- University of Toronto (CA)
- Venue
- 2020 53rd Annual IEEE/ACM International Symposium on Microarchitecture (MICRO)
- Publisher
- IEEE
- Published
- 2020-10
- Pages
- 811-824
- ISSN
- 1072-4451 (ISSN-L)
- OpenAlex
- W3022810465
- Cited by
- 10
- References
- 6
- Citations by year
- 2025: 3 · 2024: 2 · 2021: 3 · 2020: 2
- FWCI
- 0.39
- Type
- conference paper
- Language
- en
- Open access
- green
- export.arxiv.org
- Licence
- Locations
- IEEE/ACM International Symposium on Microarchitecture (MICRO) (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Engineering › Electrical and Electronic Engineering › Low-power high-performance VLSI design
- Physical Sciences › Computer Science › Hardware and Architecture › Parallel Computing and Optimization Techniques
- Keywords
- attention-based models, BERT, model quantization, 3-bit quantization, low-precision inference, memory compression, inference latency, energy efficiency, hardware co-design, TPU, tensor cores, on-chip memory
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
