Computer and Systems Architecture Commons^™

Automated Program Profiling And Analysis For Managing Heterogeneous Memory Systems, Adam Palmer Howard

Masters Theses

Many promising memory technologies, such as non-volatile, storage-class memories and high-bandwidth, on-chip RAMs, are beginning to emerge. Since each of these new technologies present tradeoffs distinct from conventional DRAMs, next-generation systems are likely to include multiple tiers of memory storage, each with their own type of devices. To efficiently utilize the available hardware, such systems will need to alter their data management strategies to consider the performance and capabilities provided by each tier.

This work explores a variety of cross-layer strategies for managing application data in heterogeneous memory systems. We propose different program profiling-based techniques to automatically partition program allocation …

Analyzing Spark Performance On Spot Instances, Jiannan Tian

Masters Theses

Amazon Spot Instances provide inexpensive service for high-performance computing. With spot instances, it is possible to get at most 90% off as discount in costs by bidding spare Amazon Elastic Computer Cloud (Amazon EC2) instances. In exchange for low cost, spot instances bring the reduced reliability onto the computing environment, because this kind of instance could be revoked abruptly by the providers due to supply and demand, and higher-priority customers are first served.

To achieve high performance on instances with compromised reliability, Spark is applied to run jobs. In this thesis, a wide set of spark experiments are conducted to …

Efficient Scaling Of A Web Proxy Cluster, Hao Zhang

Masters Theses

With the continuing growth in network traffic and increasing diversity in web content, web caching, together with various network functions (NFs), has been introduced to enhance security, optimize network performance, and save expenses. In a large enterprise network with more than tens of thousands of users, a single proxy server is not enough to handle a large number of requests and turns to group processing. When multiple web cache proxies are working as a cluster, they talk with each other and share cached objects by using internet cache protocol (ICP). This leads to poor scalability.

This thesis describes the development …

Tiled Danna: Dynamic Adaptive Neural Network Array Scaled Across Multiple Chips, Patricia Jean Eckhart

Masters Theses

Tiled Dynamic Adaptive Neural Network Array(Tiled DANNA) is a recurrent spiking neural network structure composed of programmable biologically inspired neurons and synapses that scales across multiple FPGA chips. Fire events that occur on and within DANNA initiate spiking behaviors in the programmable elements allowing DANNA to hold memory through the synaptic charge propagation and neuronal charge accumulation. DANNA is a fully digital neuromorphic computing structure based on the NIDA architecture. To support initial prototyping and testing of the Tiled DANNA, multiple Xilinx Virtex 7 690Ts were leveraged. The primary goal of Tiled DANNA is to support scaling of DANNA neural …

Scalable High-Speed Communications For Neuromorphic Systems, Aaron Reed Young

Masters Theses

Field-programmable gate arrays (FPGA), application-specific integrated circuits (ASIC), and other chip/multi-chip level implementations can be used to implement Dynamic Adaptive Neural Network Arrays (DANNA). In some applications, DANNA interfaces with a traditional computing system to provide neural network configuration information, provide network input, process network outputs, and monitor the state of the network. The present host-to-DANNA network communication setup uses a Cypress USB 3.0 peripheral controller (FX3) to enable host-to-array communication over USB 3.0. This communications setup has to run commands in batches and does not have enough bandwidth to meet the maximum throughput requirements of the DANNA device, resulting …

Optimization Of Spatial Convolution In Convnets On Intel Knl, Sangamesh Nagashattappa Ragate

Masters Theses

Most of the experts admit that the true behavior of the neural network is hard to predict. It is quite impossible to deterministically prove the working of the neural network as the architecture gets bigger, yet, it is observed that it is possible to apply a well engineered network to solve one of the most abstract problems like image recognition with substantial accuracy. It requires enormous amount of training of a considerably big and complex neural network to understand its behavior and iteratively improve its accuracy in solving a certain problem. Deep Neural Networks, which are fairly popular nowadays deal …