Paper proposes NASAIC framework for co-designing neural architectures and heterogeneous ASICs.
problem Designing efficient neural architectures and ASICs for multiple tasks.
method Build ASIC templates and propose NASAIC framework for simultaneous design of architectures and ASICs.
result NASAIC ensures design specifications and maximizes accuracy with minimal performance loss.
Improved BER with reduced power in time-domain digital backpropagation.
problem Improving BER performance in time-domain digital backpropagation.
method Jointly optimized and quantized chromatic dispersion filters using machine learning.
result Improved BER performance and power dissipation reductions.
Quantizes neural networks to eliminate multiplications, achieving high accuracy.
problem Efficiently implementing neural networks on hardware with limited resources.
method Replace activation-weight multiplication with table-based lookup for multiply-free inference.
result Achieves high accuracy with minimal multiplications, outperforming other multiply-free networks.
CirCNN compresses deep neural networks using block-circulant matrices.
problem Efficiency and accuracy trade-offs in large-scale deep neural networks.
method CirCNN uses block-circulant matrices for weight representation and processing, reducing computational and storage complexity.
result CirCNN achieves high energy efficiency and performance with negligible accuracy loss.
New KWS neural networks improve accuracy and power efficiency.
problem Maximizing accuracy and power efficiency in KWS systems.
method Hardware Aware Training (HAT) for LMU-based neural networks.
result Achieved state-of-the-art accuracy and low parameter counts.
Researchers show NN-based communication algorithms can be implemented on hardware without significant performance loss.
problem Reducing complexity and improving performance of NN-based communication algorithms for practical hardware implementation.
method Implementation of NN-based algorithms in fixed-point arithmetic with quantized weights on specialized hardware (FPGAs, ASICs).
result It is possible to implement NN-based algorithms in fixed-point arithmetic with quantized weights on hardware without significant performance loss.
CoCoPIE shows AI can run on regular devices without special hardware.
problem Real-time AI on mobile devices without special hardware.
method Compression-compiler co-design using CoCoPIE framework.
result Real-time AI applications on off-the-shelf mobile devices.
This paper improves neural network efficiency by combining filter columns and retraining, boosting array utilization and accuracy.
problem Efficient implementation of sparse convolutional neural networks on systolic arrays.
method Column combining of filter matrices, retraining of remaining weights, joint optimization for high utilization and accuracy.
result Significantly increased systolic array utilization efficiency (e.g., ~4x) and maintained high classification accuracy.
Binary and ternary weights simplify RNNs for mobile devices.
problem Complexity and memory intensity of RNNs on mobile devices.
method Learn binary and ternary weights during training.
result Significant memory saving and inference speedup on ASIC platform.
New attacks exploit neural network energy and latency, increasing costs by 10-200x.
problem Energy and latency vulnerabilities in neural networks.
method Crafted sponge examples to maximize energy and latency.
result Adversaries can increase energy consumption by 10-200x and delay decisions.
Bitcoin and other cryptocurrencies are vulnerable to quantum attacks, but some alternatives offer better security.
problem Quantum computers pose a threat to the security of cryptocurrencies like Bitcoin.
method Analysis of Bitcoin's proof-of-work and elliptic curve signature scheme, evaluation of alternative proof-of-work and post-quantum signature schemes.
result Bitcoin's proof-of-work is resistant to quantum speedup for the next 10 years, but elliptic curve signatures are vulnerable.
In January 3, 2009, Satoshi Nakamoto gave rise to the "Bitcoin Block Chain" creating the first block of the chain hashing on his computers central processing unit (CPU). Since then, the hash calculations to mine Bitcoin have been getting more and more complex, and consequently the mining hardware evolved to adapt to th…
OverQ increases model accuracy by handling outliers in neural networks with minimal hardware changes.
problem Handling outliers in neural network weights and activations for low-precision quantization.
method Overwrite quantization (OverQ) that opportunistically increases bitwidth for activation outliers.
result OverQ can handle over 90% of outliers and achieve +5% ImageNet Top-1 accuracy on a quantized ResNet-50 at 4 bits.
Channel gating reduces CNN computation cost by skipping ineffective feature regions.
problem Reducing computation cost in CNNs while maintaining accuracy.
method Dynamic, fine-grained pruning scheme that identifies and skips computation on ineffective feature regions.
result 2.7-8.0x reduction in FLOPs and 2.0-4.4x reduction in memory accesses with minimal accuracy loss.
Ternary MobileNets improve efficiency and accuracy on constrained devices.
problem Efficiently compressing MobileNets for real-time applications on constrained devices.
method Per-layer hybrid filter banks for ternary quantization of MobileNets.
result 27.98% energy savings and 51.07% reduction in model size with comparable accuracy.
Lop enables efficient machine learning models through customizable data representations and approximate computing.
problem Efficient deployment of machine learning models with reduced computational and storage complexity.
method Integrates low-precision representation and approximate arithmetic operations into machine learning frameworks and hardware.
result Lop facilitates high-level simulations and hardware realization of machine learning models with customizable data representations and approximate arithmetic operations.
This paper evaluates non-structured DNN weight pruning and finds it inferior to structured pruning.
problem Reducing energy consumption in deep neural networks.
method Developed ADMM-NN-S framework for fair comparison of non-structured and structured pruning.
result Non-structured pruning is inferior to structured pruning in terms of both storage and computation efficiency.