Look-up Table Design for Deep Sub-threshold through Full-Supply Operation

Abusultan, Monther; Khatri, Sunil P.

doi:10.1109/fccm.2014.80

Cited by 4 publications

(1 citation statement)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Namely, in this example, we use a lookup table of size b2 n , where b and n equal to 1 and 6, respectively. Table 2 compares the cost of the 6-input neuron implementation with 32-bit FP addition versus implementation with 6-input LUT in terms of energy, delay, and area [29,1]. From the table, we can see that both the delay and energy consumption of the FP addition sechme are orders of magnitude higher than that of the 6-input LUT.…”

Section: Efficiency Of Different Operationsmentioning

confidence: 99%

DTNN: Energy-efficient Inference with Dendrite Tree Inspired Neural Networks for Edge Vision Applications

Luo,

Tang,

Lee

et al. 2021

Preprint

View full text Add to dashboard Cite

Deep neural networks (DNN) have achieved remarkable success in computer vision (CV). However, training and inference of DNN models are both memory and computation intensive, incurring significant overhead in terms of energy consumption and silicon area. In particular, inference is much more cost-sensitive than training because training can be done offline with powerful platforms, while inference may have to be done on battery powered devices with constrained form factors, especially for mobile or edge vision applications. In order to accelerate DNN inference, model quantization was proposed. However previous works only focus on the quantization rate without considering the efficiency of operations. In this paper, we propose Dendrite-Tree based Neural Network (DTNN) for energy-efficient inference with table lookup operations enabled by activation quantization. In DTNN both costly weight access and arithmetic computations are eliminated for inference. We conducted experiments on various kinds of DNN models such as LeNet-5, MobileNet, VGG, and ResNet with different datasets, including MNIST, Cifar10/Cifar100, SVHN, and ImageNet. DTNN achieved significant energy saving (19.4× and 64.9× improvement on ResNet-18 and VGG-11 with ImageNet, respectively) with negligible loss of accuracy. To further validate the effectiveness of DTNN and compare with state-of-the-art low energy implementation for edge vision, we design and implement DTNN based MLP image classifiers using off-the-shelf FP-GAs. The results show that DTNN on the FPGA, with higher accuracy, could achieve orders of magnitude better energy consumption and latency compared with the state-of-the-art low energy approaches reported that use ASIC chips.

show abstract

Section: Efficiency Of Different Operationsmentioning

confidence: 99%