An efficient sparse LSTM accelerator on embedded FPGAs with bandwidth-oriented pruning

An efficient sparse LSTM accelerator on embedded FPGAs with bandwidth-oriented pruning

Long short-term memory (LSTM) networks have been widely used in natural language processing applications. Although over 80% weights can be pruned to reduce the memory requirement with little accuracy loss, the pruned model still cannot be buffered on-chip for small embedded FPGAs. Considering that w...

Saved in:

書目詳細資料
Main Authors:	Li, Shiqing, Zhu, Shien, Luo, Xiangzhong, Luo, Tao, Liu, Weichen
其他作者:	School of Computer Science and Engineering
格式:	Conference or Workshop Item
語言:	English
出版:	2023
主題:	Engineering::Computer science and engineering Engineering::Computer science and engineering::Hardware Sparse LSTM Pruning Bandwidth
在線閱讀:	https://hdl.handle.net/10356/172603
標簽:	添加標簽沒有標簽, 成為第一個標記此記錄!
機構:	Nanyang Technological University
語言:	English

相似書籍

An efficient gustavson-based sparse matrix-matrix multiplication accelerator on embedded FPGAs
由: Li, Shiqing, et al.
出版: (2023)

Accelerating sparse matrix operations on FPGAs with on/off-chip memories
由: Li, Shiqing
出版: (2023)

Accelerating gustavson-based SpMM on embedded FPGAs with element-wise parallelism and access pattern-aware caches
由: Li, Shiqing, et al.
出版: (2023)

Crossbar-aligned & integer-only neural network compression for efficient in-memory acceleration
由: Huai, Shuo, et al.
出版: (2023)

CRIMP: compact & reliable DNN inference on in-memory processing via crossbar-aligned compression and non-ideality adaptation
由: Huai, Shuo, et al.
出版: (2023)

Optimized data reuse via reordering for sparse matrix-vector multiplication on FPGAs
由: Li, Shiqing, et al.
出版: (2022)

Evaluating the merits of ranking in structured network pruning
由: Sharma, Kuldeep, et al.
出版: (2021)

A Fast Bandwidth Minimization Algorithm
由: LIM, Andrew, et al.
出版: (2007)

Single channel speech separation with constrained utterance level permutation invariant training using grid LSTM
由: Xu, Chenglin, et al.
出版: (2020)

Accelerating sparse matrix-vector multiplication on GPUs using bit-representation-optimized schemes
由: Tang, W.T., et al.
出版: (2014)

Sentinel-1 spatiotemporal simulation using convolutional LSTM for flood mapping
由: Ulloa, Noel Ivan, et al.
出版: (2022)

Towards efficient convolutional neural network for embedded hardware via multi-dimensional pruning
由: Kong, Hao, et al.
出版: (2023)

A case for energy-efficient acceleration of graph problems using embedded FPGA-based SoCs
由: Moorthy, Pradeep, et al.
出版: (2018)

Enhancing automotive embedded systems with FPGAs
由: Shanker, Shreejith
出版: (2016)

Weighting and pruning based ensemble deep random vector functional link network for tabular data classification
由: Shi, Qiushi, et al.
出版: (2023)

Dynamically-biased fixed-point LSTM for time series processing in AIoT edge device
由: Hu, Jinhai, et al.
出版: (2024)

Rethinking pruning for accelerating deep inference at the edge
由: GAO, Dawei, et al.
出版: (2020)

Pruning-aware merging for efficient multitask inference
由: GAO, Dawei, et al.
出版: (2021)

Sec71 functions as a GEF for the small GTPase Arf1 to govern dendrite pruning of Drosophila sensory neurons
由: Wang, Yan, et al.
出版: (2017)

Classification of ECG anomaly with dynamically-biased LSTM for continuous cardiac monitoring
由: Hu, Jinhai, et al.
出版: (2024)

iMAT: energy-efficient in-memory acceleration for ternary neural networks with sparse dot product
由: Zhu, Shien, et al.
出版: (2023)

Parallelizing Sparse Matrix Solve for SPICE Circuit Simulation using FPGAs
由: Kapre, Nachiket, et al.
出版: (2015)

iMAD: an in-memory accelerator for AdderNet with efficient 8-bit addition and subtraction operations
由: Zhu, Shien, et al.
出版: (2022)

Speaker and phoneme-aware speech bandwidth extension with residual dual-path network
由: Hou, Nana, et al.
出版: (2020)

EMNAPE: efficient multi-dimensional neural architecture pruning for EdgeAI
由: Kong, Hao, et al.
出版: (2023)

Alternative to extended block sparse Bayesian learning and its relation to pattern-coupled sparse Bayesian learning
由: Wang, Lu, et al.
出版: (2020)

SoftSkip: Empowering multi-modal dynamic pruning for single-stage referring comprehension
由: WEERAKOON, Dulanga, et al.
出版: (2022)

Rate-distortion optimized sparse coding with ordered dictionary for image set compression
由: Zhang, Xinfeng, et al.
出版: (2020)

Hardware-aware neural architecture search and compression towards embedded intelligence
由: Luo, Xiangzhong
出版: (2023)

Multi-task learning for end-to-end noise-robust bandwidth extension
由: Hou, Nana, et al.
出版: (2020)

Low-complexity pruning for accelerating corner detection
由: Srikanthan, Thambipillai, et al.
出版: (2013)

Structured Bayesian learning for recovery of clustered sparse signal
由: Wang, Lu, et al.
出版: (2022)

Wide-bandwidth triboelectric energy harvester combining impact nonlinearity and multi-resonance method
由: Zhao, Chaoyang, et al.
出版: (2023)

Repairing algebraic geometry codes
由: Jin, Lingfei, et al.
出版: (2020)

Pruning meta-trained networks for on-device adaptation
由: GAO, Dawei, et al.
出版: (2021)

Pruning Blocks for CNN Compression and Acceleration via Online Ensemble Distillation
由: Wang, Z., et al.
出版: (2022)

AutoPruner: transformer-based call graph pruning
由: LE, Cong Thanh, et al.
出版: (2022)

FUNCTION OF A MICROTUBULE POLYMERASE COMPLEX IN REGULATING NEURONAL PRUNING
由: TANG QUAN
出版: (2019)

Utility distribution matters: enabling fast belief propagation for multi-agent optimization with dense local utility function
由: Deng, Yanchen, et al.
出版: (2022)

Efficient FPGA-based sparse matrix-vector multiplication with data reuse-aware compression
由: Li, Shiqing, et al.
出版: (2023)