Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userBPLight-CNN: A Photonics-based Backpropagation Accelerator for Deep Learning
70
0
0.0
(
0
)
Added by
Sudeep Pasricha
Publication date
2021
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
Training deep learning networks involves continuous weight updates across the various layers of the deep network while using a backpropagation algorithm (BP). This results in expensive computation overheads during training. Consequently, most deep learning accelerators today employ pre-trained weights and focus only on improving the design of the inference phase. The recent trend is to build a complete deep learning accelerator by incorporating the training module. Such efforts require an ultra-fast chip architecture for executing the BP algorithm. In this article, we propose a novel photonics-based backpropagation accelerator for high performance deep learning training. We present the design for a convolutional neural network, BPLight-CNN, which incorporates the silicon photonics-based backpropagation accelerator. BPLight-CNN is a first-of-its-kind photonic and memristor-based CNN architecture for end-to-end training and prediction. We evaluate BPLight-CNN using a photonic CAD framework (IPKISS) on deep learning benchmark models including LeNet and VGG-Net. The proposed design achieves (i) at least 34x speedup, 34x improvement in computational efficiency, and 38.5x energy savings, during training; and (ii) 29x speedup, 31x improvement in computational efficiency, and 38.7x improvement in energy savings, during inference compared to the state-of-the-art designs. All these comparisons are done at a 16-bit resolution; and BPLight-CNN achieves these improvements at a cost of approximately 6% lower accuracy compared to the state-of-the-art.
rate research
Read More
Sparse neural networks can greatly facilitate the deployment of neural networks on resource-constrained platforms as they offer compact model sizes while retaining inference accuracy. Because of the sparsity in parameter matrices, sparse neural networks can, in principle, be exploited in accelerator architectures for improved energy-efficiency and latency. However, to realize these improvements in practice, there is a need to explore sparsity-aware hardware-software co-design. In this paper, we propose a novel silicon photonics-based sparse neural network inference accelerator called SONIC. Our experimental analysis shows that SONIC can achieve up to 5.8x better performance-per-watt and 8.4x lower energy-per-bit than state-of-the-art sparse electronic neural network accelerators; and up to 13.8x better performance-per-watt and 27.6x lower energy-per-bit than the best known photonic neural network accelerators.
The increasing application of deep learning technology drives the need for an efficient parallel computing architecture for Convolutional Neural Networks (CNNs). A significant challenge faced when designing a many-core CNN accelerator is to handle the data movement between the processing elements. The CNN workload introduces many-to-one traffic in addition to one-to-one and one-to-many traffic. As the de-facto standard for on-chip communication, Network-on-Chip (NoC) can support various unicast and multicast traffic. For many-to-one traffic, repetitive unicast is employed which is not an efficient way. In this paper, we propose to use the gather packet on mesh-based NoCs employing output stationary systolic array in support of many-to-one traffic. The gather packet will collect the data from the intermediate nodes eventually leading to the destination efficiently. This method is evaluated using the traffic traces generated from the convolution layer of AlexNet and VGG-16 with improvement in the latency and power than the repetitive unicast method.
The recent breakthroughs and prohibitive complexities of Deep Neural Networks (DNNs) have excited extensive interest in domain-specific DNN accelerators, among which optical DNN accelerators are particularly promising thanks to their unprecedented potential of achieving superior performance-per-watt. However, the development of optical DNN accelerators is much slower than that of electrical DNN accelerators. One key challenge is that while many techniques have been developed to facilitate the development of electrical DNN accelerators, techniques that support or expedite optical DNN accelerator design remain much less explored, limiting both the achievable performance and the innovation development of optical DNN accelerators. To this end, we develop the first-of-its-kind framework dubbed O-HAS, which for the first time demonstrates automated Optical Hardware Accelerator Search for boosting both the acceleration efficiency and development speed of optical DNN accelerators. Specifically, our O-HAS consists of two integrated enablers: (1) an O-Cost Predictor, which can accurately yet efficiently predict an optical accelerators energy and latency based on the DNN model parameters and the optical accelerator design; and (2) an O-Search Engine, which can automatically explore the large design space of optical DNN accelerators and identify the optimal accelerators (i.e., the micro-architectures and algorithm-to-accelerator mapping methods) in order to maximize the target acceleration efficiency. Extensive experiments and ablation studies consistently validate the effectiveness of both our O-Cost Predictor and O-Search Engine as well as the excellent efficiency of O-HAS generated optical accelerators.
Memristor crossbars are circuits capable of performing analog matrix-vector multiplications, overcoming the fundamental energy efficiency limitations of digital logic. They have been shown to be effective in special-purpose accelerators for a limited set of neural network applications. We present the Programmable Ultra-efficient Memristor-based Accelerator (PUMA) which enhances memristor crossbars with general purpose execution units to enable the acceleration of a wide variety of Machine Learning (ML) inference workloads. PUMAs microarchitecture techniques exposed through a specialized Instruction Set Architecture (ISA) retain the efficiency of in-memory computing and analog circuitry, without compromising programmability. We also present the PUMA compiler which translates high-level code to PUMA ISA. The compiler partitions the computational graph and optimizes instruction scheduling and register allocation to generate code for large and complex workloads to run on thousands of spatial cores. We have developed a detailed architecture simulator that incorporates the functionality, timing, and power models of PUMAs components to evaluate performance and energy consumption. A PUMA accelerator running at 1 GHz can reach area and power efficiency of $577~GOPS/s/mm^2$ and $837~GOPS/s/W$, respectively. Our evaluation of diverse ML applications from image recognition, machine translation, and language modelling (5M-800M synapses) shows that PUMA achieves up to $2,446times$ energy and $66times$ latency improvement for inference compared to state-of-the-art GPUs. Compared to an application-specific memristor-based accelerator, PUMA incurs small energy overheads at similar inference latency and added programmability.
In this paper, we propose a very compact embedded CNN processor design based on a modified logarithmic computing method using very low bit-width representation. Our high-quality CNN processor can easily fit into edge devices. For Yolov2, our processing circuit takes only 0.15 mm2 using TSMC 40 nm cell library. The key idea is to constrain the activation and weight values of all layers uniformly to be within the range [-1, 1] and produce low bit-width logarithmic representation. With the uniform representations, we devise a unified, reusable CNN computing kernel and significantly reduce computing resources. The proposed approach has been extensively evaluated on many popular image classification CNN models (AlexNet, VGG16, and ResNet-18/34) and object detection models (Yolov2). The hardware-implemented results show that our design consumes only minimal computing and storage resources, yet attains very high accuracy. The design is thoroughly verified on FPGAs, and the SoC integration is underway with promising results. With extremely efficient resource and energy usage, our design is excellent for edge computing purposes.
suggested questions
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
