arXiv subjects

Damitha Lenadora

Publications and source records attributed to Damitha Lenadora.

6 recordsLinked to original sources

arXiv2025

COGNATE: Acceleration of Sparse Tensor Programs on Emerging Hardware using Transfer Learning

Sparse tensor programs are essential in deep learning and graph analytics, driving the need for optimized processing. To meet this demand, specialized hardware accelerators are being developed. Optimizing these programs for accelerators is challenging for two reasons: program performance is highly sensitive to variations in sparse inputs, and early-stage accelerators rely on expensive simulators. Therefore, ML-based cost models used for optimizing such programs on general-purpose hardware are often ineffective for early-stage accelerators, as they require large datasets for proper training. To this end, we introduce COGNATE, a novel framework that leverages inexpensive data samples from general-purpose hardware (e.g., CPUs) to train cost models, followed by few-shot fine-tuning on emerging hardware. COGNATE exploits the homogeneity of input features across hardware platforms while effectively mitigating heterogeneity, enabling cost model training with just 5% of the data samples needed by accelerator-specific models to achieve comparable performance. We conduct extensive experiments to demonstrate that COGNATE outperforms existing techniques, achieving average speedups of 1.47x (up to 5.46x) for SpMM and 1.39x (up to 4.22x) for SDDMM.

cs.LG↗

arXiv2023

SENSEi: Input-Sensitive Compilation for Accelerating GNNs

Over the years, many frameworks and optimization techniques have been proposed to accelerate graph neural networks (GNNs). Compared to the optimizations explored in these systems, we observe that different matrix re-associations of GNN computations lead to novel input-sensitive performance behavior. We leverage this observation to propose SENSEi, a system that exposes different sparse and dense matrix primitive compositions based on different matrix re-associations of GNN computations and selects the best among them based on input attributes. SENSEi executes in two stages: (1) an offline compilation stage that enumerates all valid re-associations leading to different sparse-dense matrix compositions and uses input-oblivious pruning techniques to prune away clearly unprofitable candidates and (2) an online runtime system that explores the remaining candidates and uses light-weight cost models to select the best re-association based on the input graph and the embedding sizes on a given hardware platform. On a wide range of configurations, SENSEi achieves speedups of up to $2.012\times$ and $1.85\times$ on graph convolutional networks and up to $6.294\times$ and $16.274\times$ on graph attention networks, on GPUs and CPUs respectively. We also show that its technique generalizes to GNN variants, including those that require sampling. Furthermore, we show that SENSEi's techniques are agnostic to the underlying GNN system, and can be used to yield synergistic improvements across a diverse set of implementations.

cs.LG↗

arXiv2022

Hybrid Cloud and HPC Approach to High-Performance Dataframes

Data pre-processing is a fundamental component in any data-driven application. With the increasing complexity of data processing operations and volume of data, Cylon, a distributed dataframe system, is developed to facilitate data processing both as a standalone application and as a library, especially for Python applications. While Cylon shows promising performance results, we experienced difficulties trying to integrate with frameworks incompatible with the traditional Message Passing Interface (MPI). While MPI implementations encompass scalable and efficient communication routines, their process launching mechanisms work well with mainstream HPC systems but are incompatible with some environments that adopt their own resource management systems. In this work, we alleviated this issue by directly integrating the Unified Communication X (UCX) framework, which supports a variety of classic HPC and non-HPC process-bootstrapping mechanisms as our communication framework. While we experimented with our methodology on Cylon, the same technique can be used to bring MPI communication to other applications that do not employ MPI's built-in process management approach.

cs.DC↗

arXiv2022

High Performance Dataframes from Parallel Processing Patterns

The data science community today has embraced the concept of Dataframes as the de facto standard for data representation and manipulation. Ease of use, massive operator coverage, and popularization of R and Python languages have heavily influenced this transformation. However, most widely used serial Dataframes today (R, pandas) experience performance limitations even while working on even moderately large data sets. We believe that there is plenty of room for improvement by investigating the generic distributed patterns of dataframe operators. In this paper, we propose a framework that lays the foundation for building high performance distributed-memory parallel dataframe systems based on these parallel processing patterns. We also present Cylon, as a reference runtime implementation. We demonstrate how this framework has enabled Cylon achieving scalable high performance. We also underline the flexibility of the proposed API and the extensibility of the framework on different hardware. To the best of our knowledge, Cylon is the first and only distributed-memory parallel dataframe system available today.

cs.DC↗

arXiv2020

A Fast, Scalable, Universal Approach For Distributed Data Aggregations

In the current era of Big Data, data engineering has transformed into an essential field of study across many branches of science. Advancements in Artificial Intelligence (AI) have broadened the scope of data engineering and opened up new applications in both enterprise and research communities. Aggregations (also termed reduce in functional programming) are an integral functionality in these applications. They are traditionally aimed at generating meaningful information on large data-sets, and today, they are being used for engineering more effective features for complex AI models. Aggregations are usually carried out on top of data abstractions such as tables/ arrays and are combined with other operations such as grouping of values. There are frameworks that excel in the said domains individually. But, we believe that there is an essential requirement for a data analytics tool that can universally integrate with existing frameworks, and thereby increase the productivity and efficiency of the entire data analytics pipeline. Cylon endeavors to fulfill this void. In this paper, we present Cylon's fast and scalable aggregation operations implemented on top of a distributed in-memory table structure that universally integrates with existing frameworks.

cs.DC↗

arXiv2020

Exploratory Analysis of a Social Media Network in Sri Lanka during the COVID-19 Virus Outbreak

During the COVID-19 pandemic, multiple aspects of human life were subjected to unprecedented changes, globally. In Sri Lanka, a developing country located in South Asia, it was possible to observe a range of events that arose due to the influence of the COVID-19 virus outbreak. Thus, the people of Sri Lanka used Social Media to voice their opinions regarding such events and those involved in them, enabling the ideal avenue to explore the social perception. However, the outcome of such actions was at certain times detrimental. This study was conducted as an attempt to identify the reasons for such instances as well as to identify the behaviours of the Sri Lankan populace during such a crisis event. To support this study, observations, as well as data of related posts from a sample of 50 sources, were manually collected from the most popular social media platform in Sri Lanka, Facebook. The posts considered spanned until approximately a month after the initial major virus outbreak in the country and contained content that even vaguely related to the virus. Utilising such data, various forms of analyses such as topic significance and topic co-occurrences were conducted. The findings highlight, while there can be social detrimental ideas shared, the majority of the posts point constructive and positive thoughts suggesting the successful influence from the cultural and social values Sri Lanka society promotes throughout.

cs.SI↗