Lookupx: Next-Generation Quantization and Lookup Techniques for Empowering Performance and Energy Efficiency

Cagla Irmak Rumelili Koksal*, Nihat Mert Cicek, Ayse Yilmazer Metin, Berna Ors

*Bu çalışma için yazışmadan sorumlu yazar

Araştırma sonucu: ???type-name???Konferans katkısıbilirkişi

Özet

Long Short Term Memory (LSTM) networks as one of the most used Recurrent Neural Networks (RNN) structures offer high accuracy for sequence learning tasks. However, it is challenging to offer low latency and high throughput while satisfying the low power constraints at the same time for computationally expensive LSTM operations. This work offers a two-pronged approach to accelerate inference in RNN networks. First, linear quantization technique is applied to reduce the complexity of operations, power consumption and required memory resources. Then, a new activation implementation method is proposed, called lookupx, to accelerate sigmoid function computation during inference. It is shown that lowering precision to 4-bit integer numbers for inputs causes only 2% accuracy loss and the lookupx activation methodology has 1.9x better performance and 50x lower power consumption while decreasing the required chip area 1.2x compared to integer domain activation functions with the same accuracy result.

Orijinal dilİngilizce
Ana bilgisayar yayını başlığıICECS 2023 - 2023 30th IEEE International Conference on Electronics, Circuits and Systems
Ana bilgisayar yayını alt yazısıTechnosapiens for Saving Humanity
YayınlayanInstitute of Electrical and Electronics Engineers Inc.
ISBN (Elektronik)9798350326499
DOI'lar
Yayın durumuYayınlandı - 2023
Etkinlik30th IEEE International Conference on Electronics, Circuits and Systems, ICECS 2023 - Istanbul, Turkey
Süre: 4 Ara 20237 Ara 2023

Yayın serisi

AdıICECS 2023 - 2023 30th IEEE International Conference on Electronics, Circuits and Systems: Technosapiens for Saving Humanity

???event.eventtypes.event.conference???

???event.eventtypes.event.conference???30th IEEE International Conference on Electronics, Circuits and Systems, ICECS 2023
Ülke/BölgeTurkey
ŞehirIstanbul
Periyot4/12/237/12/23

Bibliyografik not

Publisher Copyright:
© 2023 IEEE.

Parmak izi

Lookupx: Next-Generation Quantization and Lookup Techniques for Empowering Performance and Energy Efficiency' araştırma başlıklarına git. Birlikte benzersiz bir parmak izi oluştururlar.

Alıntı Yap