Data-Aware Scheduling in Datacenters

Datacenters have emerged as the dominant form of computing infrastructure over the last two decades. The tremendous increase in the requirements of data analysis has led to a proportional increase in power consumption and datacenters are now one of the fastest growing electricity consumers in the United States. Another rising concern is the loss of throughput due to network congestion. Scheduling models that do not explicitly account for data placement may lead to a transfer of large amounts of data over the network causing unacceptable delays. In this dissertation, we study different scheduling models that are inspired by the dual objectives of minimizing energy costs and network congestion in a datacenter. As datacenters are equipped to handle peak workloads, the average server utilization in most datacenters is very low. As a result, one can achieve huge energy savings by selectively shutting down machines when demand is low. In this dissertation, we introduce the network-aware machine activation problem to find a schedule that simultaneously minimizes the number of machines necessary and the congestion incurred in the network. Our model significantly generalizes well-studied combinatorial optimization problems such as hard-capacitated hypergraph covering and is thus strongly NP-hard. As a result, we focus on finding good approximation algorithms. Data-parallel computation frameworks such as MapReduce have popularized the design of applications that require a large amount of communication between different machines. Efficient scheduling of these communication demands is essential to guarantee efficient execution of the different applications. In the second part of the thesis, we study the approximability of the co-flow scheduling problem that has been recently introduced to capture these application-level demands. Finally, we also study the question, "In what order should one process jobs?'' Often, precedence constraints specify a partial order over the set of jobs and the objective is to find suitable schedules that satisfy the partial order. However, in the presence of hard deadline constraints, it may be impossible to find a schedule that satisfies all precedence constraints. In this thesis we formalize different variants of job scheduling with soft precedence constraints and conduct the first systematic study of these problems.

Multi-level, Multi-stage and Stochastic Optimization Models for Energy Conservation in Buildings for Federal, State and Local Agencies

Energy Conservation Measure (ECM) project selection is made difficult given real-world constraints, limited resources to implement savings retrofits, various suppliers in the market and project financing alternatives. Many of these energy efficient retrofit projects should be viewed as a series of investments with annual returns for these traditionally risk-averse agencies. Given a list of ECMs available, federal, state and local agencies must determine how to implement projects at lowest costs. The most common methods of implementation planning are suboptimal relative to cost. Federal, state and local agencies can obtain greater returns on their energy conservation investment over traditional methods, regardless of the implementing organization. This dissertation outlines several approaches to improve the traditional energy conservations models. Any public buildings in regions with similar energy conservation goals in the United States or internationally can also benefit greatly from this research. Additionally, many private owners of buildings are under mandates to conserve energy e.g., Local Law 85 of the New York City Energy Conservation Code requires any building, public or private, to meet the most current energy code for any alteration or renovation. Thus, both public and private stakeholders can benefit from this research. The research in this dissertation advances and presents models that decision-makers can use to optimize the selection of ECM projects with respect to the total cost of implementation. A practical application of a two-level mathematical program with equilibrium constraints (MPEC) improves the current best practice for agencies concerned with making the most cost-effective selection leveraging energy services companies or utilities. The two-level model maximizes savings to the agency and profit to the energy services companies (Chapter 2). An additional model presented leverages a single congressional appropriation to implement ECM projects (Chapter 3). Returns from implemented ECM projects are used to fund additional ECM projects. In these cases, fluctuations in energy costs and uncertainty in the estimated savings severely influence ECM project selection and the amount of the appropriation requested. A risk aversion method proposed imposes a minimum on the number of “of projects completed in each stage. A comparative method using Conditional Value at Risk is analyzed. Time consistency was addressed in this chapter. This work demonstrates how a risk-based, stochastic, multi-stage model with binary decision variables at each stage provides a much more accurate estimate for planning than the agency’s traditional approach and deterministic models. Finally, in Chapter 4, a rolling-horizon model allows for subadditivity and superadditivity of the energy savings to simulate interactive effects between ECM projects. The approach makes use of inequalities (McCormick, 1976) to re-express constraints that involve the product of binary variables with an exact linearization (related to the convex hull of those constraints). This model additionally shows the benefits of learning between stages while remaining consistent with the single congressional appropriations framework.

Architectural-Physical Co-Design of 3D CPUs with Micro-Fluidic Cooling

The performance, energy efficiency and cost improvements due to traditional technology scaling have begun to slow down and present diminishing returns. Underlying reasons for this trend include fundamental physical limits of transistor scaling, the growing significance of quantum effects as transistors shrink, and a growing mismatch between transistors and interconnects regarding size, speed and power. Continued Moore's Law scaling will not come from technology scaling alone, and must involve improvements to design tools and development of new disruptive technologies such as 3D integration. 3D integration presents potential improvements to interconnect power and delay by translating the routing problem into a third dimension, and facilitates transistor density scaling independent of technology node. Furthermore, 3D IC technology opens up a new architectural design space of heterogeneously-integrated high-bandwidth CPUs. Vertical integration promises to provide the CPU architectures of the future by integrating high performance processors with on-chip high-bandwidth memory systems and highly connected network-on-chip structures. Such techniques can overcome the well-known CPU performance bottlenecks referred to as memory and communication wall. However the promising improvements to performance and energy efficiency offered by 3D CPUs does not come without cost, both in the financial investments to develop the technology, and the increased complexity of design. Two main limitations to 3D IC technology have been heat removal and TSV reliability. Transistor stacking creates increases in power density, current density and thermal resistance in air cooled packages. Furthermore the technology introduces vertical through silicon vias (TSVs) that create new points of failure in the chip and require development of new BEOL technologies. Although these issues can be controlled to some extent using thermal-reliability aware physical and architectural 3D design techniques, high performance embedded cooling schemes, such as micro-fluidic (MF) cooling, are fundamentally necessary to unlock the true potential of 3D ICs. A new paradigm is being put forth which integrates the computational, electrical, physical, thermal and reliability views of a system. The unification of these diverse aspects of integrated circuits is called Co-Design. Independent design and optimization of each aspect leads to sub-optimal designs due to a lack of understanding of cross-domain interactions and their impacts on the feasibility region of the architectural design space. Co-Design enables optimization across layers with a multi-domain view and thus unlocks new high-performance and energy efficient configurations. Although the co-design paradigm is becoming increasingly necessary in all fields of IC design, it is even more critical in 3D ICs where, as we show, the inter-layer coupling and higher degree of connectivity between components exacerbates the interdependence between architectural parameters, physical design parameters and the multitude of metrics of interest to the designer (i.e. power, performance, temperature and reliability). In this dissertation we present a framework for multi-domain co-simulation and co-optimization of 3D CPU architectures with both air and MF cooling solutions. Finally we propose an approach for design space exploration and modeling within the new Co-Design paradigm, and discuss the possible avenues for improvement of this work in the future.

RICH AND EFFICIENT VISUAL DATA REPRESENTATION

Increasing the size of training data in many computer vision tasks has shown to be very effective. Using large scale image datasets (e.g. ImageNet) with simple learning techniques (e.g. linear classifiers) one can achieve state-of-the-art performance in object recognition compared to sophisticated learning techniques on smaller image sets. Semantic search on visual data has become very popular. There are billions of images on the internet and the number is increasing every day. Dealing with large scale image sets is intense per se. They take a significant amount of memory that makes it impossible to process the images with complex algorithms on single CPU machines. Finding an efficient image representation can be a key to attack this problem. A representation being efficient is not enough for image understanding. It should be comprehensive and rich in carrying semantic information. In this proposal we develop an approach to computing binary codes that provide a rich and efficient image representation. We demonstrate several tasks in which binary features can be very effective. We show how binary features can speed up large scale image classification. We present learning techniques to learn the binary features from supervised image set (With different types of semantic supervision; class labels, textual descriptions). We propose several problems that are very important in finding and using efficient image representation.

Learning Binary Code Representations for Effective and Efficient Image Retrieval

The size of online image datasets is constantly increasing. Considering an image dataset with millions of images, image retrieval becomes a seemingly intractable problem for exhaustive similarity search algorithms. Hashing methods, which encodes high-dimensional descriptors into compact binary strings, have become very popular because of their high efficiency in search and storage capacity. In the first part, we propose a multimodal retrieval method based on latent feature models. The procedure consists of a nonparametric Bayesian framework for learning underlying semantically meaningful abstract features in a multimodal dataset, a probabilistic retrieval model that allows cross-modal queries and an extension model for relevance feedback. In the second part, we focus on supervised hashing with kernels. We describe a flexible hashing procedure that treats binary codes and pairwise semantic similarity as latent and observed variables, respectively, in a probabilistic model based on Gaussian processes for binary classification. We present a scalable inference algorithm with the sparse pseudo-input Gaussian process (SPGP) model and distributed computing. In the last part, we define an incremental hashing strategy for dynamic databases where new images are added to the databases frequently. The method is based on a two-stage classification framework using binary and multi-class SVMs. The proposed method also enforces balance in binary codes by an imbalance penalty to obtain higher quality binary codes. We learn hash functions by an efficient algorithm where the NP-hard problem of finding optimal binary codes is solved via cyclic coordinate descent and SVMs are trained in a parallelized incremental manner. For modifications like adding images from an unseen class, we propose an incremental procedure for effective and efficient updates to the previous hash functions. Experiments on three large-scale image datasets demonstrate that the incremental strategy is capable of efficiently updating hash functions to the same retrieval performance as hashing from scratch.

ESSAYS IN ENERGY, ENVIRONMENT AND TECHNOLOGICAL CHANGE

This dissertation studies technological change in the context of energy and environmental economics. Technology plays a key role in reducing greenhouse gas emissions from the transportation sector. Chapter 1 estimates a structural model of the car industry that allows for endogenous product characteristics to investigate how gasoline taxes, R&D subsidies and competition affect fuel efficiency and vehicle prices in the medium-run, both through car-makers' decisions to adopt technologies and through their investments in knowledge capital. I use technology adoption and automotive patents data for 1986-2006 to estimate this model. I show that 92% of fuel efficiency improvements between 1986 and 2006 were driven by technology adoption, while the role of knowledge capital is largely to reduce the marginal production costs of fuel-efficient cars. A counterfactual predicts that an additional $1/gallon gasoline tax in 2006 would have increased the technology adoption rate, and raised average fuel efficiency by 0.47 miles/gallon, twice the annual fuel efficiency improvement in 2003-2006. An R&D subsidy that would reduce the marginal cost of knowledge capital by 25% in 2006 would have raised investment in knowledge capital. This subsidy would have raised fuel efficiency only by 0.06 miles/gallon in 2006, but would have increased variable profits by $2.3 billion over all firms that year. Passenger vehicle fuel economy standards in the United States will require substantial improvements in new vehicle fuel economy over the next decade. Economic theory suggests that vehicle manufacturers adopt greater fuel-saving technologies for vehicles with larger market size. Chapter 2 documents a strong connection between market size, measured by sales, and technology adoption. Using variation consumer demographics and purchasing pattern to account for the endogeneity of market size, we find that a 10 percent increase in market size raises vehicle fuel efficiency by 0.3 percent, as compared to a mean improvement of 1.4 percent per year over 1997-2013. Historically, fuel price and demographic-driven market size changes have had large effects on technology adoption. Furthermore, fuel taxes would induce firms to adopt fuel-saving technologies on their most efficient cars, thereby polarizing the fuel efficiency distribution of the new vehicle fleet.

Design and Pilot Study for an Efficient High-Throughput Automated Computer-Vision Guided Intelligent De-Calyxing Machine for Post-Harvest Strawberry Processing

Strawberries harvested for processing as frozen fruits are currently de-calyxed manually in the field. This process requires the removal of the stem cap with green leaves (i.e. the calyx) and incurs many disadvantages when performed by hand. Not only does it necessitate the need to maintain cutting tool sanitation, but it also increases labor time and exposure of the de-capped strawberries before in-plant processing. This leads to labor inefficiency and decreased harvest yield. By moving the calyx removal process from the fields to the processing plants, this new practice would reduce field labor and improve management and logistics, while increasing annual yield. As labor prices continue to increase, the strawberry industry has shown great interest in the development and implementation of an automated calyx removal system. In response, this dissertation describes the design, operation, and performance of a full-scale automatic vision-guided intelligent de-calyxing (AVID) prototype machine. The AVID machine utilizes commercially available equipment to produce a relatively low cost automated de-calyxing system that can be retrofitted into existing food processing facilities. This dissertation is broken up into five sections. The first two sections include a machine overview and a 12-week processing plant pilot study. Results of the pilot study indicate the AVID machine is able to de-calyx grade-1-with-cap conical strawberries at roughly 66 percent output weight yield at a throughput of 10,000 pounds per hour. The remaining three sections describe in detail the three main components of the machine: a strawberry loading and orientation conveyor, a machine vision system for calyx identification, and a synchronized multi-waterjet knife calyx removal system. In short, the loading system utilizes rotational energy to orient conical strawberries. The machine vision system determines cut locations through RGB real-time feature extraction. The high-speed multi-waterjet knife system uses direct drive actuation to locate 30,000 psi cutting streams to precise coordinates for calyx removal. Based on the observations and studies performed within this dissertation, the AVID machine is seen to be a viable option for automated high-throughput strawberry calyx removal. A summary of future tasks and further improvements is discussed at the end.

the plenoptic sensor

In this thesis, we will introduce the innovative concept of a plenoptic sensor that can determine the phase and amplitude distortion in a coherent beam, for example a laser beam that has propagated through the turbulent atmosphere.. The plenoptic sensor can be applied to situations involving strong or deep atmospheric turbulence. This can improve free space optical communications by maintaining optical links more intelligently and efficiently. Also, in directed energy applications, the plenoptic sensor and its fast reconstruction algorithm can give instantaneous instructions to an adaptive optics (AO) system to create intelligent corrections in directing a beam through atmospheric turbulence. The hardware structure of the plenoptic sensor uses an objective lens and a microlens array (MLA) to form a mini “Keplerian” telescope array that shares the common objective lens. In principle, the objective lens helps to detect the phase gradient of the distorted laser beam and the microlens array (MLA) helps to retrieve the geometry of the distorted beam in various gradient segments. The software layer of the plenoptic sensor is developed based on different applications. Intuitively, since the device maximizes the observation of the light field in front of the sensor, different algorithms can be developed, such as detecting the atmospheric turbulence effects as well as retrieving undistorted images of distant objects. Efficient 3D simulations on atmospheric turbulence based on geometric optics have been established to help us perform optimization on system design and verify the correctness of our algorithms. A number of experimental platforms have been built to implement the plenoptic sensor in various application concepts and show its improvements when compared with traditional wavefront sensors. As a result, the plenoptic sensor brings a revolution to the study of atmospheric turbulence and generates new approaches to handle turbulence effect better.