Thomas Parnell
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Thomas Parnell has a total of 28 co-authors. Here are the top co-authors by number of shared publications:
- Haralampos Pozidis
IBM Zurich Research Laboratory
4 shared publications - Andreea Anghel
IBM Research – Zurich, Rüschlikon, Switzerland
2 shared publications - Celestine Dünner
IBM Research Zurich SWITZERLAND
2 shared publications Dimitrios Sarigiannis
IBM Research
2 shared publicationsNikolaos Papandreou
2 shared publications- Nikolas Ioannou
IBM Research Europe IBM Research Europe IBM Research Europe Independent researcher
2 shared publications
Latest Publications
- Open Access
LLM-Pilot: Characterize and Optimize Performance of your LLM Inference Services
As Large Language Models (LLMs) are rapidly growing in popularity, LLM inference services must be able to serve requests from thousands of users while satisfying performance requirements. The performance of an LLM inference service is largely determined by the hardware onto which it is deployed, but understanding of which hardware will deliver on performance requirements remains challenging. In this work we present LLM-Pilot - a first-of-its-kind system for characterizing and predicting performance of LLM inference services. LLM-Pilot performs benchmarking of LLM inference services, under a realistic workload, across a variety of GPUs, and optimizes the service configuration for each considered GPU to maximize performance. Finally, using this characterization data, LLM-Pilot learns a predictive model, which can be used to recommend the most cost-effective hardware for a previously unseen LLM. Compared to existing methods, LLM-Pilot can deliver on performance requirements 33% more frequently, whilst reducing costs by 60% on average.
Open AccessLLM-Pilot: Characterize and Optimize Performance of your LLM Inference Services
As Large Language Models (LLMs) are rapidly growing in popularity, LLM inference services must be able to serve requests from thousands of users while satisfying performance requirements. The performance of an LLM inference service is largely determined by the hardware onto which it is deployed, but understanding of which hardware will deliver on performance requirements remains challenging. In this work we present LLM-Pilot - a first-of-its-kind system for characterizing and predicting performance of LLM inference services. LLM-Pilot performs benchmarking of LLM inference services, under a realistic workload, across a variety of GPUs, and optimizes the service configuration for each considered GPU to maximize performance. Finally, using this characterization data, LLM-Pilot learns a predictive model, which can be used to recommend the most cost-effective hardware for a previously unseen LLM. Compared to existing methods, LLM-Pilot can deliver on performance requirements 33% more frequently, whilst reducing costs by 60% on average.
- 14 April 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 70 (6) , 922-935
Efficient Pipelined Execution of CNNs Based on In-Memory Computing and Graph Homomorphism Verification
In-memory computing is an emerging computing paradigm enabling deep-learning inference at significantly higher energy-efficiency and reduced latency. The essential idea is mapping the synaptic weights of each layer to one or more in-memory computing (ICM) cores. During inference, these cores perform the associated matrix-vector multiplications in place with O(1) time complexity, obviating the need to move the synaptic weights to additional processing units. Moreover, this architecture enables the execution of these networks in a highly pipelined fashion. However, a key challenge is designing an efficient communication fabric for the CM cores. In this work, we present one such communication fabric based on a graph topology that is well-suited for the widely successful convolutional neural networks (CNNs). We show that this communication fabric facilitates the pipelined execution of all state-of-the-art CNNs by proving the existence of a homomorphism between the graph representations of these networks and that corresponding to the proposed communication fabric. We then present a quantitative comparison with established communication topologies and show that our proposed topology achieves the lowest bandwidth requirements per communication channel. Finally, we present one hardware implementation and show a concrete example of mapping ResNet-32 onto an CM cores array interconnected via the proposed communication fabric.
- 14 April 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 70 (6) , 922-935
Efficient Pipelined Execution of CNNs Based on In-Memory Computing and Graph Homomorphism Verification
In-memory computing is an emerging computing paradigm enabling deep-learning inference at significantly higher energy-efficiency and reduced latency. The essential idea is mapping the synaptic weights of each layer to one or more in-memory computing (ICM) cores. During inference, these cores perform the associated matrix-vector multiplications in place with O(1) time complexity, obviating the need to move the synaptic weights to additional processing units. Moreover, this architecture enables the execution of these networks in a highly pipelined fashion. However, a key challenge is designing an efficient communication fabric for the CM cores. In this work, we present one such communication fabric based on a graph topology that is well-suited for the widely successful convolutional neural networks (CNNs). We show that this communication fabric facilitates the pipelined execution of all state-of-the-art CNNs by proving the existence of a homomorphism between the graph representations of these networks and that corresponding to the proposed communication fabric. We then present a quantitative comparison with established communication topologies and show that our proposed topology achieves the lowest bandwidth requirements per communication channel. Finally, we present one hardware implementation and show a concrete example of mapping ResNet-32 onto an CM cores array interconnected via the proposed communication fabric.
- Open Access
SnapBoost: A Heterogeneous Boosting Machine
Modern gradient boosting software frameworks, such as XGBoost and LightGBM, implement Newton descent in a functional space. At each boosting iteration, their goal is to find the base hypothesis, selected from some base hypothesis class, that is closest to the Newton descent direction in a Euclidean sense. Typically, the base hypothesis class is fixed to be all binary decision trees up to a given depth. In this work, we study a Heterogeneous Newton Boosting Machine (HNBM) in which the base hypothesis class may vary across boosting iterations. Specifically, at each boosting iteration, the base hypothesis class is chosen, from a fixed set of subclasses, by sampling from a probability distribution. We derive a global linear convergence rate for the HNBM under certain assumptions, and show that it agrees with existing rates for Newton's method when the Newton direction can be perfectly fitted by the base hypothesis at each boosting iteration. We then describe a particular realization of a HNBM, SnapBoost, that, at each boosting iteration, randomly selects between either a decision tree of variable depth or a linear regressor with random Fourier features. We describe how SnapBoost is implemented, with a focus on the training complexity. Finally, we present experimental results, using OpenML and Kaggle datasets, that show that SnapBoost is able to achieve better generalization loss than competing boosting frameworks, without taking significantly longer to tune.
Open AccessSnapBoost: A Heterogeneous Boosting Machine
Modern gradient boosting software frameworks, such as XGBoost and LightGBM, implement Newton descent in a functional space. At each boosting iteration, their goal is to find the base hypothesis, selected from some base hypothesis class, that is closest to the Newton descent direction in a Euclidean sense. Typically, the base hypothesis class is fixed to be all binary decision trees up to a given depth. In this work, we study a Heterogeneous Newton Boosting Machine (HNBM) in which the base hypothesis class may vary across boosting iterations. Specifically, at each boosting iteration, the base hypothesis class is chosen, from a fixed set of subclasses, by sampling from a probability distribution. We derive a global linear convergence rate for the HNBM under certain assumptions, and show that it agrees with existing rates for Newton's method when the Newton direction can be perfectly fitted by the base hypothesis at each boosting iteration. We then describe a particular realization of a HNBM, SnapBoost, that, at each boosting iteration, randomly selects between either a decision tree of variable depth or a linear regressor with random Fourier features. We describe how SnapBoost is implemented, with a focus on the training complexity. Finally, we present experimental results, using OpenML and Kaggle datasets, that show that SnapBoost is able to achieve better generalization loss than competing boosting frameworks, without taking significantly longer to tune.
- Open Access
Learning to Tune XGBoost with XGBoost
In this short paper we investigate whether meta-learning techniques can be used to more effectively tune the hyperparameters of machine learning models using successive halving (SH). We propose a novel variant of the SH algorithm (MeSH), that uses meta-regressors to determine which candidate configurations should be eliminated at each round. We apply MeSH to the problem of tuning the hyperparameters of a gradient-boosted decision tree model. By training and tuning our meta-regressors using existing tuning jobs from 95 datasets, we demonstrate that MeSH can often find a superior solution to both SH and random search.
Open AccessLearning to Tune XGBoost with XGBoost
In this short paper we investigate whether meta-learning techniques can be used to more effectively tune the hyperparameters of machine learning models using successive halving (SH). We propose a novel variant of the SH algorithm (MeSH), that uses meta-regressors to determine which candidate configurations should be eliminated at each round. We apply MeSH to the problem of tuning the hyperparameters of a gradient-boosted decision tree model. By training and tuning our meta-regressors using existing tuning jobs from 95 datasets, we demonstrate that MeSH can often find a superior solution to both SH and random search.
- Open Access
Weighted Sampling for Combined Model Selection and Hyperparameter Tuning
The combined algorithm selection and hyperparameter tuning (CASH) problem is characterized by large hierarchical hyperparameter spaces. Model-free hyperparameter tuning methods can explore such large spaces efficiently since they are highly parallelizable across multiple machines. When no prior knowledge or meta-data exists to boost their performance, these methods commonly sample random configurations following a uniform distribution. In this work, we propose a novel sampling distribution as an alternative to uniform sampling and prove theoretically that it has a better chance of finding the best configuration in a worst-case setting. In order to compare competing methods rigorously in an experimental setting, one must perform statistical hypothesis testing. We show that there is little-to-no agreement in the automated machine learning literature regarding which methods should be used. We contrast this disparity with the methods recommended by the broader statistics literature, and identify a suitable approach. We then select three popular model-free solutions to CASH and evaluate their performance, with uniform sampling as well as the proposed sampling scheme, across 67 datasets from the OpenML platform. We investigate the trade-off between exploration and exploitation across the three algorithms, and verify empirically that the proposed sampling distribution improves performance in all cases.
Open AccessWeighted Sampling for Combined Model Selection and Hyperparameter Tuning
The combined algorithm selection and hyperparameter tuning (CASH) problem is characterized by large hierarchical hyperparameter spaces. Model-free hyperparameter tuning methods can explore such large spaces efficiently since they are highly parallelizable across multiple machines. When no prior knowledge or meta-data exists to boost their performance, these methods commonly sample random configurations following a uniform distribution. In this work, we propose a novel sampling distribution as an alternative to uniform sampling and prove theoretically that it has a better chance of finding the best configuration in a worst-case setting. In order to compare competing methods rigorously in an experimental setting, one must perform statistical hypothesis testing. We show that there is little-to-no agreement in the automated machine learning literature regarding which methods should be used. We contrast this disparity with the methods recommended by the broader statistics literature, and identify a suitable approach. We then select three popular model-free solutions to CASH and evaluate their performance, with uniform sampling as well as the proposed sampling scheme, across 67 datasets from the OpenML platform. We investigate the trade-off between exploration and exploitation across the three algorithms, and verify empirically that the proposed sampling distribution improves performance in all cases.
- Open Access
SySCD: A System-Aware Parallel Coordinate Descent Algorithm
In this paper we propose a novel parallel stochastic coordinate descent (SCD) algorithm with convergence guarantees that exhibits strong scalability. We start by studying a state-of-the-art parallel implementation of SCD and identify scalability as well as system-level performance bottlenecks of the respective implementation. We then take a principled approach to develop a new SCD variant which is designed to avoid the identified system bottlenecks, such as limited scaling due to coherence traffic of model sharing across threads, and inefficient CPU cache accesses. Our proposed system-aware parallel coordinate descent algorithm (SySCD) scales to many cores and across numa nodes, and offers a consistent bottom line speedup in training time of up to x12 compared to an optimized asynchronous parallel SCD algorithm and up to x42, compared to state-of-the-art GLM solvers (scikit-learn, Vowpal Wabbit, and H2O) on a range of datasets and multi-core CPU architectures.
Open AccessSySCD: A System-Aware Parallel Coordinate Descent Algorithm
In this paper we propose a novel parallel stochastic coordinate descent (SCD) algorithm with convergence guarantees that exhibits strong scalability. We start by studying a state-of-the-art parallel implementation of SCD and identify scalability as well as system-level performance bottlenecks of the respective implementation. We then take a principled approach to develop a new SCD variant which is designed to avoid the identified system bottlenecks, such as limited scaling due to coherence traffic of model sharing across threads, and inefficient CPU cache accesses. Our proposed system-aware parallel coordinate descent algorithm (SySCD) scales to many cores and across numa nodes, and offers a consistent bottom line speedup in training time of up to x12 compared to an optimized asynchronous parallel SCD algorithm and up to x42, compared to state-of-the-art GLM solvers (scikit-learn, Vowpal Wabbit, and H2O) on a range of datasets and multi-core CPU architectures.
- 1 October 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Reliability of 3D NAND flash memory with a focus on read voltage calibration from a system aspect
This paper discusses the reliability challenges of 3D NAND flash memory and their impact on flash management for enterprise storage applications. Emphasis is given to the read voltage calibration and its critical role in achieving low error-rates and low latency read performance, as well as in enabling accurate block health estimation. We present experimental results that demonstrate the improvements in endurance, retention and read-disturb from different read voltage calibration schemes, and we address their requirements from a system perspective, i.e., the accuracy vs. complexity trade-off. We discuss the above aspects for state-of-the-art 3D TLC and QLC NAND flash memory.
- 1 October 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
Reliability of 3D NAND flash memory with a focus on read voltage calibration from a system aspect
This paper discusses the reliability challenges of 3D NAND flash memory and their impact on flash management for enterprise storage applications. Emphasis is given to the read voltage calibration and its critical role in achieving low error-rates and low latency read performance, as well as in enabling accurate block health estimation. We present experimental results that demonstrate the improvements in endurance, retention and read-disturb from different read voltage calibration schemes, and we address their requirements from a system perspective, i.e., the accuracy vs. complexity trade-off. We discuss the above aspects for state-of-the-art 3D TLC and QLC NAND flash memory.
- Open Access
Understanding and Optimizing the Performance of Distributed Machine Learning Applications on Apache Spark
In this paper we explore the performance limits of Apache Spark for machine learning applications. We begin by analyzing the characteristics of a state-of-the-art distributed machine learning algorithm implemented in Spark and compare it to an equivalent reference implementation using the high performance computing framework MPI. We identify critical bottlenecks of the Spark framework and carefully study their implications on the performance of the algorithm. In order to improve Spark performance we then propose a number of practical techniques to alleviate some of its overheads. However, optimizing computational efficiency and framework related overheads is not the only key to performance -- we demonstrate that in order to get the best performance out of any implementation it is necessary to carefully tune the algorithm to the respective trade-off between computation time and communication latency. The optimal trade-off depends on both the properties of the distributed algorithm as well as infrastructure and framework-related characteristics. Finally, we apply these technical and algorithmic optimizations to three different distributed linear machine learning algorithms that have been implemented in Spark. We present results using five large datasets and demonstrate that by using the proposed optimizations, we can achieve a reduction in the performance difference between Spark and MPI from 20x to 2x.
Open AccessUnderstanding and Optimizing the Performance of Distributed Machine Learning Applications on Apache Spark
In this paper we explore the performance limits of Apache Spark for machine learning applications. We begin by analyzing the characteristics of a state-of-the-art distributed machine learning algorithm implemented in Spark and compare it to an equivalent reference implementation using the high performance computing framework MPI. We identify critical bottlenecks of the Spark framework and carefully study their implications on the performance of the algorithm. In order to improve Spark performance we then propose a number of practical techniques to alleviate some of its overheads. However, optimizing computational efficiency and framework related overheads is not the only key to performance -- we demonstrate that in order to get the best performance out of any implementation it is necessary to carefully tune the algorithm to the respective trade-off between computation time and communication latency. The optimal trade-off depends on both the properties of the distributed algorithm as well as infrastructure and framework-related characteristics. Finally, we apply these technical and algorithmic optimizations to three different distributed linear machine learning algorithms that have been implemented in Spark. We present results using five large datasets and demonstrate that by using the proposed optimizations, we can achieve a reduction in the performance difference between Spark and MPI from 20x to 2x.
- Open Access
Efficient Use of Limited-Memory Accelerators for Linear Learning on Heterogeneous Systems
We propose a generic algorithmic building block to accelerate training of machine learning models on heterogeneous compute systems. Our scheme allows to efficiently employ compute accelerators such as GPUs and FPGAs for the training of large-scale machine learning models, when the training data exceeds their memory capacity. Also, it provides adaptivity to any system's memory hierarchy in terms of size and processing speed. Our technique is built upon novel theoretical insights regarding primal-dual coordinate methods, and uses duality gap information to dynamically decide which part of the data should be made available for fast processing. To illustrate the power of our approach we demonstrate its performance for training of generalized linear models on a large-scale dataset exceeding the memory size of a modern GPU, showing an order-of-magnitude speedup over existing approaches.
Open AccessEfficient Use of Limited-Memory Accelerators for Linear Learning on Heterogeneous Systems
We propose a generic algorithmic building block to accelerate training of machine learning models on heterogeneous compute systems. Our scheme allows to efficiently employ compute accelerators such as GPUs and FPGAs for the training of large-scale machine learning models, when the training data exceeds their memory capacity. Also, it provides adaptivity to any system's memory hierarchy in terms of size and processing speed. Our technique is built upon novel theoretical insights regarding primal-dual coordinate methods, and uses duality gap information to dynamically decide which part of the data should be made available for fast processing. To illustrate the power of our approach we demonstrate its performance for training of generalized linear models on a large-scale dataset exceeding the memory size of a modern GPU, showing an order-of-magnitude speedup over existing approaches.
- 26 August 2016
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Selected Areas in Communications
- Vol. 34 (9) , 2354-2365
Capacity of the MLC NAND Flash Channel
In this paper, we develop a framework for evaluating the symmetric capacity of multilevel-cell (MLC) NAND flash devices while making very few assumptions regarding the underlying device physics. A set of recursive equations are derived that allow one to measure the symmetric capacity for any given page in a flash device using simple conditional statistics that can be extracted experimentally. Using data captured from two different 1y nm MLC devices, we demonstrate that the symmetric capacity of a flash page not only depends on the amount of program/erase cycling and data retention stress that has accumulated, but also on the position of the page within the flash block. We then study the effect on symmetric capacity of using optimized read-back schemes (both hard and soft) and show that while there is significant benefit, not all pages in the block are improved by the same amount. Finally, we show that it is possible to design error correction architectures that harness the inherent variation of symmetric capacity within a flash block to dramatically extend the program/erase cycling endurance of flash-based storage systems.
- 26 August 2016
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal on Selected Areas in Communications
- Vol. 34 (9) , 2354-2365
Capacity of the MLC NAND Flash Channel
In this paper, we develop a framework for evaluating the symmetric capacity of multilevel-cell (MLC) NAND flash devices while making very few assumptions regarding the underlying device physics. A set of recursive equations are derived that allow one to measure the symmetric capacity for any given page in a flash device using simple conditional statistics that can be extracted experimentally. Using data captured from two different 1y nm MLC devices, we demonstrate that the symmetric capacity of a flash page not only depends on the amount of program/erase cycling and data retention stress that has accumulated, but also on the position of the page within the flash block. We then study the effect on symmetric capacity of using optimized read-back schemes (both hard and soft) and show that while there is significant benefit, not all pages in the block are improved by the same amount. Finally, we show that it is possible to design error correction architectures that harness the inherent variation of symmetric capacity within a flash block to dramatically extend the program/erase cycling endurance of flash-based storage systems.