arXiv subjects

Henry Hoffmann

Publications and source records attributed to Henry Hoffmann.

26 records · Page 2Linked to original sources

arXiv2022

Cello: Efficient Computer Systems Optimization with Predictive Early Termination and Censored Regression

Sample-efficient machine learning (SEML) has been widely applied to find optimal latency and power tradeoffs for configurable computer systems. Instead of randomly sampling from the configuration space, SEML reduces the search cost by dramatically reducing the number of configurations that must be sampled to optimize system goals (e.g., low latency or energy). Nevertheless, SEML only reduces one component of cost -- the total number of samples collected -- but does not decrease the cost of collecting each sample. Critically, not all samples are equal; some take much longer to collect because they correspond to slow system configurations. This paper present Cello, a computer systems optimization framework that reduces sample collection costs -- especially those that come from the slowest configurations. The key insight is to predict ahead of time whether samples will have poor system behavior (e.g., long latency or high energy) and terminate these samples early before their measured system behavior surpasses the termination threshold, which we call it predictive early termination. To predict the future system behavior accurately before it manifests as high runtime or energy, Cello uses censored regression to produces accurate predictions for running samples. We evaluate Cello by optimizing latency and energy for Apache Spark workloads. We give Cello a fixed amount of time to search a combined space of hardware and software configuration parameters. Our evaluation shows that compared to the state-of-the-art SEML approach in computer systems optimization, Cello improves latency by 1.19X for minimizing latency under a power constraint, and improves energy by 1.18X for minimizing energy under a latency constraint.

cs.LG↗

arXiv2020

Orthogonalized SGD and Nested Architectures for Anytime Neural Networks

We propose a novel variant of SGD customized for training network architectures that support anytime behavior: such networks produce a series of increasingly accurate outputs over time. Efficient architectural designs for these networks focus on re-using internal state; subnetworks must produce representations relevant for both immediate prediction as well as refinement by subsequent network stages. We consider traditional branched networks as well as a new class of recursively nested networks. Our new optimizer, Orthogonalized SGD, dynamically re-balances task-specific gradients when training a multitask network. In the context of anytime architectures, this optimizer projects gradients from later outputs onto a parameter subspace that does not interfere with those from earlier outputs. Experiments demonstrate that training with Orthogonalized SGD significantly improves generalization accuracy of anytime networks.

cs.LG↗

arXiv2020

ALERT: Accurate Learning for Energy and Timeliness

An increasing number of software applications incorporate runtime Deep Neural Networks (DNNs) to process sensor data and return inference results to humans. Effective deployment of DNNs in these interactive scenarios requires meeting latency and accuracy constraints while minimizing energy, a problem exacerbated by common system dynamics. Prior approaches handle dynamics through either (1) system-oblivious DNN adaptation, which adjusts DNN latency/accuracy tradeoffs, or (2) application-oblivious system adaptation, which adjusts resources to change latency/energy tradeoffs. In contrast, this paper improves on the state-of-the-art by coordinating application- and system-level adaptation. ALERT, our runtime scheduler, uses a probabilistic model to detect environmental volatility and then simultaneously select both a DNN and a system resource configuration to meet latency, accuracy, and energy constraints. We evaluate ALERT on CPU and GPU platforms for image and speech tasks in dynamic environments. ALERT's holistic approach achieves more than 13% energy reduction, and 27% error reduction over prior approaches that adapt solely at the application or system level. Furthermore, ALERT incurs only 3% more energy consumption and 2% higher DNN-inference error than an oracle scheme with perfect application and system knowledge.

cs.PF↗

arXiv2020

QURATOR: Innovative Technologies for Content and Data Curation

In all domains and sectors, the demand for intelligent systems to support the processing and generation of digital content is rapidly increasing. The availability of vast amounts of content and the pressure to publish new content quickly and in rapid succession requires faster, more efficient and smarter processing and generation methods. With a consortium of ten partners from research and industry and a broad range of expertise in AI, Machine Learning and Language Technologies, the QURATOR project, funded by the German Federal Ministry of Education and Research, develops a sustainable and innovative technology platform that provides services to support knowledge workers in various industries to address the challenges they face when curating digital content. The project's vision and ambition is to establish an ecosystem for content curation technologies that significantly pushes the current state of the art and transforms its region, the metropolitan area Berlin-Brandenburg, into a global centre of excellence for curation technologies.

cs.DL↗

arXiv2019

Partial Compilation of Variational Algorithms for Noisy Intermediate-Scale Quantum Machines

Quantum computing is on the cusp of reality with Noisy Intermediate-Scale Quantum (NISQ) machines currently under development and testing. Some of the most promising algorithms for these machines are variational algorithms that employ classical optimization coupled with quantum hardware to evaluate the quality of each candidate solution. Recent work used GRadient Descent Pulse Engineering (GRAPE) to translate quantum programs into highly optimized machine control pulses, resulting in a significant reduction in the execution time of programs. This is critical, as quantum machines can barely support the execution of short programs before failing. However, GRAPE suffers from high compilation latency, which is untenable in variational algorithms since compilation is interleaved with computation. We propose two strategies for partial compilation, exploiting the structure of variational circuits to pre-compile optimal pulses for specific blocks of gates. Our results indicate significant pulse speedups ranging from 1.5x-3x in typical benchmarks, with only a small fraction of the compilation latency of GRAPE.

quant-ph↗

arXiv2019

Strain controllable magnetocrystalline anisotropy in FeRh/MgO bilayers

Ultra-thin film of FeRh on insulator MgO substrate has been investigated usingab-initio electronic structure calculations. From this calculation, we have found the interesting effect of epitaxial strain on the magnetocrystalline anisotropy (MCA). Analysis of the energy and k-resolved distribution of the orbital character of the band structure reveals that MCA largely arises from the spin-orbit coupling (SOC) between dx2-y2 andd xz,yz orbitals of Fe atoms at the FeRh/MgO interface. We demonstrate that the strain has significant effects on the MCA: It not only affects the value of the MCA but also induces a switching of the magnetic easy axis from perpendicular to in-plane direction. The mechanism is the strain-induced shifts of the SOC d-states. Our work demonstrates that strain engineering can open a viable pathway towards tailoring magnetic properties for antiferromagetic spintronic applications.

cond-mat.mtrl-sci↗

arXiv2019

Language Support for Adaptation: Intent-Driven Programming in FAST

Historically, programming language semantics has focused on assigning a precise mathematical meaning to programs. That meaning is a function from the program's input domain to its output domain determined solely by its syntactic structure. Such a semantics, fosters the development of portable applications which are oblivious to the performance characteristics and limitations (such as a maximum memory footprint) of particular hardware and software platforms. This paper introduces the idea of intent-driven programming where the meaning of a program additionally depends on an accompanying intent specification expressing how the ordinary program meaning is dynamically modified during execution to satisfy additional properties expressed by the intent. These include both intensional properties---e.g., resource usage---and extensional properties---e.g., accuracy of the computed answer. To demonstrate the intent-driven programming model's value, this paper presents a general-purpose intent-driven programming language---called FAST---implemented as an extension of Swift. FAST consists of an intent compiler, a profiler, a general controller interface and a runtime module which supports interoperation with legacy C/C++ codes. Compared to existing frameworks for adaptive computing, \FAST{} supports dynamic adaptation to changes both in the operating environment and in the intent itself, and enables the mixing of procedural control and control based on feedback and optimization.

cs.PL↗

arXiv2017

Understanding and Auto-Adjusting Performance-Related Configurations

Modern software systems are often equipped with hundreds to thousands of configuration options, many of which greatly affect performance. Unfortunately, properly setting these configurations is challenging for developers due to the complex and dynamic nature of system workload and environment. In this paper, we first conduct an empirical study to understand performance-related configurations and the challenges of setting them in the real-world. Guided by our study, we design a systematic and general control-theoretic framework, SmartConf, to automatically set and dynamically adjust performance-related configurations to meet required operating constraints while optimizing other performance metrics. Evaluation shows that SmartConf is effective in solving real-world configuration problems, often providing better performance than even the best static configuration developers can choose under existing configuration systems.

cs.DC↗