Matching Items (39)

158769-Thumbnail Image.png

Efficient and Online Deep Learning through Model Plasticity and Stability

Description

The rapid advancement of Deep Neural Networks (DNNs), computing, and sensing technology has enabled many new applications, such as the self-driving vehicle, the surveillance drone, and the robotic system. Compared

The rapid advancement of Deep Neural Networks (DNNs), computing, and sensing technology has enabled many new applications, such as the self-driving vehicle, the surveillance drone, and the robotic system. Compared to conventional edge devices (e.g. cell phone or smart home devices), these emerging devices are required to deal with much more complicated and dynamic situations in real-time with bounded computation resources. However, there are several challenges, including but not limited to efficiency, real-time adaptation, model stability, and automation of architecture design.

To tackle the challenges mentioned above, model plasticity and stability are leveraged to achieve efficient and online deep learning, especially in the scenario of learning streaming data at the edge:

First, a dynamic training scheme named Continuous Growth and Pruning (CGaP) is proposed to compress the DNNs through growing important parameters and pruning unimportant ones, achieving up to 98.1% reduction in the number of parameters.

Second, this dissertation presents Progressive Segmented Training (PST), which targets catastrophic forgetting problems in continual learning through importance sampling, model segmentation, and memory-assisted balancing. PST achieves state-of-the-art accuracy with 1.5X FLOPs reduction in the complete inference path.

Third, to facilitate online learning in real applications, acquisitive learning (AL) is further proposed to emphasize both knowledge inheritance and acquisition: the majority of the knowledge is first pre-trained in the inherited model and then adapted to acquire new knowledge. The inherited model's stability is monitored by noise injection and the landscape of the loss function, while the acquisition is realized by importance sampling and model segmentation. Compared to a conventional scheme, AL reduces accuracy drop by >10X on CIFAR-100 dataset, with 5X reduction in latency per training image and 150X reduction in training FLOPs.

Finally, this dissertation presents evolutionary neural architecture search in light of model stability (ENAS-S). ENAS-S uses a novel fitness score, which addresses not only the accuracy but also the model stability, to search for an optimal inherited model for the application of continual learning. ENAS-S outperforms hand-designed DNNs when learning from a data stream at the edge.

In summary, in this dissertation, several algorithms exploiting model plasticity and model stability are presented to improve the efficiency and accuracy of deep neural networks, especially for the scenario of continual learning.

Contributors

Agent

Created

Date Created
  • 2020

152800-Thumbnail Image.png

Emerging neural coincidences in rats agranular medial and agranular lateral cortices during learning of a directional choice task

Description

To uncover the neural correlates to go-directed behavior, single unit action potentials are considered fundamental computing units and have been examined by different analytical methodologies under a broad set of

To uncover the neural correlates to go-directed behavior, single unit action potentials are considered fundamental computing units and have been examined by different analytical methodologies under a broad set of hypotheses. Using a behaving rat performing a directional choice learning task, we aim to study changes in rat's cortical neural patterns while he improved his task performance accuracy from chance to 80% or higher. Specifically, simultaneous multi-channel single unit neural recordings from the rat's agranular medial (AGm) and Agranular lateral (AGl) cortices were analyzed using joint peristimulus time histogram (JPSTHs), which effectively unveils firing coincidences in neural action potentials. My results based on data from six rats revealed that coincidences of pair-wise neural action potentials are higher when rats were performing the task than they were not at the learning stage, and this trend abated after the rats learned the task. Another finding is that the coincidences at the learning stage are stronger than that when the rats learned the task especially when they were performing the task. Therefore, this coincidence measure is the highest when the rats were performing the task at the learning stage. This may suggest that neural coincidences play a role in the coordination and communication among populations of neurons engaged in a purposeful act. Additionally, attention and working memory may have contributed to the modulation of neural coincidences during the designed task.

Contributors

Agent

Created

Date Created
  • 2014

155708-Thumbnail Image.png

6T-SRAM 1Mb design with test structures and post silicon validation

Description

Static random-access memories (SRAM) are integral part of design systems as caches and data memories that and occupy one-third of design space. The work presents an embedded low power SRAM

Static random-access memories (SRAM) are integral part of design systems as caches and data memories that and occupy one-third of design space. The work presents an embedded low power SRAM on a triple well process that allows body-biasing control. In addition to the normal mode operation, the design is embedded with Physical Unclonable Function (PUF) [Suh07] and Sense Amplifier Test (SA Test) mode. With PUF mode structures, the fabrication and environmental mismatches in bit cells are used to generate unique identification bits. These bits are fixed and known as preferred state of an SRAM bit cell. The direct access test structure is a measurement unit for offset voltage analysis of sense amplifiers. These designs are manufactured using a foundry bulk CMOS 55 nm low-power (LP) process. The details about SRAM bit-cell and peripheral circuit design is discussed in detail, for certain cases the circuit simulation analysis is performed with random variations embedded in SPICE models. Further, post-silicon testing results are discussed for normal operation of SRAMs and the special test modes. The silicon and circuit simulation results for various tests are presented.

Contributors

Agent

Created

Date Created
  • 2017

155141-Thumbnail Image.png

A mixed signal adaptive ripple cancellation technique for integrated buck converters

Description

Switching regulator has several advantages over linear regulator, but the drawback of switching regulator is ripple voltage on output. Previously people use LDO following a buck converter and multi-phase buck

Switching regulator has several advantages over linear regulator, but the drawback of switching regulator is ripple voltage on output. Previously people use LDO following a buck converter and multi-phase buck converter to reduce the output voltage ripple. However, these two solutions also have obvious drawbacks and limitations.

In this thesis, a novel mixed signal adaptive ripple cancellation technique is presented. The idea is to generate an artificial ripple current with the same amplitude as inductor current ripple but opposite phase that has high linearity tracking behavior. To generate the artificial triangular current, duty cycle information and inductor current ripple amplitude information are needed. By sensing switching node SW, the duty cycle information can be obtained; by using feedback the amplitude of the artificial ripple current can be regulated. The artificial ripple current cancels out the inductor current, and results in a very low ripple output current flowing to load. In top level simulation, 19.3dB ripple rejection can be achieved.

Contributors

Agent

Created

Date Created
  • 2016

154977-Thumbnail Image.png

The feasibility of domain specific compilation for spatially programmable architectures

Description

Integrated circuits must be energy efficient. This efficiency affects all aspects of chip design, from the battery life of embedded devices to thermal heating on high performance servers. As technology

Integrated circuits must be energy efficient. This efficiency affects all aspects of chip design, from the battery life of embedded devices to thermal heating on high performance servers. As technology scaling slows, future generations of transistors will lack the energy efficiency gains as it has had in previous generations. Therefore, other sources of energy efficiency will be much more important. Many computations have the potential to be executed for extreme energy efficiency but are not instigated because the platforms they run on are not optimized for efficient execution. ASICs improve energy efficiency by reducing flexibility and leveraging the properties of a specific computation. However, ASICs are fixed in function and therefore have incredible opportunity cost. FPGAs offer a reconfigurable solution but are 25x less energy efficient than ASIC implementation. Spatially programmable architectures (SPAs) are similar in design and structure to ASICs and FPGAs but are able bridge the ASIC-FPGA energy efficiency gap by trading flexibility for efficiency. However, SPAs are difficult to program because they do not share the same programming model as normal architectures that execute in time. This work addresses compiler challenges for coarse grained, locally interconnected SPA for domain efficiency (SPADE). A novel SPADE topology, called the wave pipeline, is introduced that is designed for the image signal processing domain that is both efficient and simple to compile to. A compiler for the wave pipeline is created that solves for maximum energy and area efficiency using low complexity, greedy methods. The wave pipeline topology and compiler allow for us to investigate and experiment with image signal processing applications to prove the feasibility of SPADE compilers.

Contributors

Agent

Created

Date Created
  • 2016

155918-Thumbnail Image.png

Accelerated Aging in Devices and Circuits

Description

The aging mechanism in devices is prone to uncertainties due to dynamic stress conditions. In AMS circuits these can lead to momentary fluctuations in circuit voltage that may be missed

The aging mechanism in devices is prone to uncertainties due to dynamic stress conditions. In AMS circuits these can lead to momentary fluctuations in circuit voltage that may be missed by a compact model and hence cause unpredictable failure. Firstly, multiple aging effects in the devices may have underlying correlations. The generation of new traps during TDDB may significantly accelerate BTI, since these traps are close to the dielectric-Si interface in scaled technology. Secondly, the prevalent reliability analysis lacks a direct validation of the lifetime of devices and circuits. The aging mechanism of BTI causes gradual degradation of the device leading to threshold voltage shift and increasing the failure rate. In the 28nm HKMG technology, contribution of BTI to NMOS degradation has become significant at high temperature as compared to Channel Hot Carrier (CHC). This requires revising the End of Lifetime (EOL) calculation based on contribution from induvial aging effects especially in feedback loops. Conventionally, aging in devices is extrapolated from a short-term measurement, but this practice results in unreliable prediction of EOL caused by variability in initial parameters and stress conditions. To mitigate the extrapolation issues and improve predictability, this work aims at providing a new approach to test the device to EOL in a fast and controllable manner. The contributions of this thesis include: (1) based on stochastic trapping/de-trapping mechanism, new compact BTI models are developed and verified with 14nm FinFET and 28nm HKMG data. Moreover, these models are implemented into circuit simulation, illustrating a significant increase in failure rate due to accelerated BTI, (2) developing a model to predict accelerated aging under special conditions like feedback loops and stacked inverters, (3) introducing a feedback loop based test methodology called Adaptive Accelerated Aging (AAA) that can generate accurate aging data till EOL, (4) presenting simulation and experimental data for the models and providing test setup for multiple stress conditions, including those for achieving EOL in 1 hour device as well as ring oscillator (RO) circuit for validation of the proposed methodology, and (5) scaling these models for finding a guard band for VLSI design circuits that can provide realistic aging impact.

Contributors

Agent

Created

Date Created
  • 2017

156845-Thumbnail Image.png

Hardware Acceleration of Deep Convolutional Neural Networks on FPGA

Description

The rapid improvement in computation capability has made deep convolutional neural networks (CNNs) a great success in recent years on many computer vision tasks with significantly improved accuracy. During the

The rapid improvement in computation capability has made deep convolutional neural networks (CNNs) a great success in recent years on many computer vision tasks with significantly improved accuracy. During the inference phase, many applications demand low latency processing of one image with strict power consumption requirement, which reduces the efficiency of GPU and other general-purpose platform, bringing opportunities for specific acceleration hardware, e.g. FPGA, by customizing the digital circuit specific for the deep learning algorithm inference. However, deploying CNNs on portable and embedded systems is still challenging due to large data volume, intensive computation, varying algorithm structures, and frequent memory accesses. This dissertation proposes a complete design methodology and framework to accelerate the inference process of various CNN algorithms on FPGA hardware with high performance, efficiency and flexibility.

As convolution contributes most operations in CNNs, the convolution acceleration scheme significantly affects the efficiency and performance of a hardware CNN accelerator. Convolution involves multiply and accumulate (MAC) operations with four levels of loops. Without fully studying the convolution loop optimization before the hardware design phase, the resulting accelerator can hardly exploit the data reuse and manage data movement efficiently. This work overcomes these barriers by quantitatively analyzing and optimizing the design objectives (e.g. memory access) of the CNN accelerator based on multiple design variables. An efficient dataflow and hardware architecture of CNN acceleration are proposed to minimize the data communication while maximizing the resource utilization to achieve high performance.

Although great performance and efficiency can be achieved by customizing the FPGA hardware for each CNN model, significant efforts and expertise are required leading to long development time, which makes it difficult to catch up with the rapid development of CNN algorithms. In this work, we present an RTL-level CNN compiler that automatically generates customized FPGA hardware for the inference tasks of various CNNs, in order to enable high-level fast prototyping of CNNs from software to FPGA and still keep the benefits of low-level hardware optimization. First, a general-purpose library of RTL modules is developed to model different operations at each layer. The integration and dataflow of physical modules are predefined in the top-level system template and reconfigured during compilation for a given CNN algorithm. The runtime control of layer-by-layer sequential computation is managed by the proposed execution schedule so that even highly irregular and complex network topology, e.g. GoogLeNet and ResNet, can be compiled. The proposed methodology is demonstrated with various CNN algorithms, e.g. NiN, VGG, GoogLeNet and ResNet, on two different standalone FPGAs achieving state-of-the art performance.

Based on the optimized acceleration strategy, there are still a lot of design options, e.g. the degree and dimension of computation parallelism, the size of on-chip buffers, and the external memory bandwidth, which impact the utilization of computation resources and data communication efficiency, and finally affect the performance and energy consumption of the accelerator. The large design space of the accelerator makes it impractical to explore the optimal design choice during the real implementation phase. Therefore, a performance model is proposed in this work to quantitatively estimate the accelerator performance and resource utilization. By this means, the performance bottleneck and design bound can be identified and the optimal design option can be explored early in the design phase.

Contributors

Agent

Created

Date Created
  • 2018

157619-Thumbnail Image.png

On-chip learning and inference acceleration of sparse representations

Description

The past decade has seen a tremendous surge in running machine learning (ML) functions on mobile devices, from mere novelty applications to now indispensable features for the next generation of

The past decade has seen a tremendous surge in running machine learning (ML) functions on mobile devices, from mere novelty applications to now indispensable features for the next generation of devices.

While the mobile platform capabilities range widely, long battery life and reliability are common design concerns that are crucial to remain competitive.

Consequently, state-of-the-art mobile platforms have become highly heterogeneous by combining a powerful CPUs with GPUs to accelerate the computation of deep neural networks (DNNs), which are the most common structures to perform ML operations.

But traditional von Neumann architectures are not optimized for the high memory bandwidth and massively parallel computation demands required by DNNs.

Hence, propelling research into non-von Neumann architectures to support the demands of DNNs.

The re-imagining of computer architectures to perform efficient DNN computations requires focusing on the prohibitive demands presented by DNNs and alleviating them. The two central challenges for efficient computation are (1) large memory storage and movement due to weights of the DNN and (2) massively parallel multiplications to compute the DNN output.

Introducing sparsity into the DNNs, where certain percentage of either the weights or the outputs of the DNN are zero, greatly helps with both challenges. This along with algorithm-hardware co-design to compress the DNNs is demonstrated to provide efficient solutions to greatly reduce the power consumption of hardware that compute DNNs. Additionally, exploring emerging technologies such as non-volatile memories and 3-D stacking of silicon in conjunction with algorithm-hardware co-design architectures will pave the way for the next generation of mobile devices.

Towards the objectives stated above, our specific contributions include (a) an architecture based on resistive crosspoint array that can update all values stored and compute matrix vector multiplication in parallel within a single cycle, (b) a framework of training DNNs with a block-wise sparsity to drastically reduce memory storage and total number of computations required to compute the output of DNNs, (c) the exploration of hardware implementations of sparse DNNs and architectural guidelines to reduce power consumption for the implementations in monolithic 3D integrated circuits, and (d) a prototype chip in 65nm CMOS accelerator for long-short term memory networks trained with the proposed block-wise sparsity scheme.

Contributors

Agent

Created

Date Created
  • 2019

154195-Thumbnail Image.png

Energy-efficient digital circuit design using threshold logic gates

Description

Improving energy efficiency has always been the prime objective of the custom and automated digital circuit design techniques. As a result, a multitude of methods to reduce power without sacrificing

Improving energy efficiency has always been the prime objective of the custom and automated digital circuit design techniques. As a result, a multitude of methods to reduce power without sacrificing performance have been proposed. However, as the field of design automation has matured over the last few decades, there have been no new automated design techniques, that can provide considerable improvements in circuit power, leakage and area. Although emerging nano-devices are expected to replace the existing MOSFET devices, they are far from being as mature as semiconductor devices and their full potential and promises are many years away from being practical.

The research described in this dissertation consists of four main parts. First is a new circuit architecture of a differential threshold logic flipflop called PNAND. The PNAND gate is an edge-triggered multi-input sequential cell whose next state function is a threshold function of its inputs. Second a new approach, called hybridization, that replaces flipflops and parts of their logic cones with PNAND cells is described. The resulting \hybrid circuit, which consists of conventional logic cells and PNANDs, is shown to have significantly less power consumption, smaller area, less standby power and less power variation.

Third, a new architecture of a field programmable array, called field programmable threshold logic array (FPTLA), in which the standard lookup table (LUT) is replaced by a PNAND is described. The FPTLA is shown to have as much as 50% lower energy-delay product compared to conventional FPGA using well known FPGA modeling tool called VPR.

Fourth, a novel clock skewing technique that makes use of the completion detection feature of the differential mode flipflops is described. This clock skewing method improves the area and power of the ASIC circuits by increasing slack on timing paths. An additional advantage of this method is the elimination of hold time violation on given short paths.

Several circuit design methodologies such as retiming and asynchronous circuit design can use the proposed threshold logic gate effectively. Therefore, the use of threshold logic flipflops in conventional design methodologies opens new avenues of research towards more energy-efficient circuits.

Contributors

Agent

Created

Date Created
  • 2015

155631-Thumbnail Image.png

Hardware Acceleration of Most Apparent Distortion Image Quality Assessment Algorithm on FPGA Using OpenCL

Description

The information era has brought about many technological advancements in the past

few decades, and that has led to an exponential increase in the creation of digital images and

videos. Constantly, all

The information era has brought about many technological advancements in the past

few decades, and that has led to an exponential increase in the creation of digital images and

videos. Constantly, all digital images go through some image processing algorithm for

various reasons like compression, transmission, storage, etc. There is data loss during this

process which leaves us with a degraded image. Hence, to ensure minimal degradation of

images, the requirement for quality assessment has become mandatory. Image Quality

Assessment (IQA) has been researched and developed over the last several decades to

predict the quality score in a manner that agrees with human judgments of quality. Modern

image quality assessment (IQA) algorithms are quite effective at prediction accuracy, and

their development has not focused on improving computational performance. The existing

serial implementation requires a relatively large run-time on the order of seconds for a single

frame. Hardware acceleration using Field programmable gate arrays (FPGAs) provides

reconfigurable computing fabric that can be tailored for a broad range of applications.

Usually, programming FPGAs has required expertise in hardware descriptive languages

(HDLs) or high-level synthesis (HLS) tool. OpenCL is an open standard for cross-platform,

parallel programming of heterogeneous systems along with Altera OpenCL SDK, enabling

developers to use FPGA's potential without extensive hardware knowledge. Hence, this

thesis focuses on accelerating the computationally intensive part of the most apparent

distortion (MAD) algorithm on FPGA using OpenCL. The results are compared with CPU

implementation to evaluate performance and efficiency gains.

Contributors

Agent

Created

Date Created
  • 2017