Open Source Artifacts
Besides peer-reviewed research publications, our group also makes our contributions available as open-source artifacts. Although not required by the publication venues, we make these artifacts available at the time of publication of our regular peer-reviewed research publications. The goals behind making these contributions open-source are threefold: to expedite research progress as a community, lower the barrier to entry to new fields, and promote research reproducibility.
Lowcarb: Reinforcement Learning-based Carbon-Aware Scheduling of Serverless Functions
A Reinforcement Learning-based carbon-aware scheduling framework with for serving serverless functions.
WaterSplit: Water-Aware On-Site and Off-Site Water Allocation Framework
A water-aware optimization framework that balances on-site cooling water and off-site electricity-related water in datacenters.
GreenMix: Randomized Sketching Based Energy-Efficient Serverless Computing Framework
A framework for energy-efficient serverless function execution on asymmetric multi-core processors.
Ribbon: Cost-Effective and QoS-Aware Deep Learning Model Inference System
A cost-effective and QoS-aware system for deep learning model inference.
Systematically Inferring I/O Performance Variability by Examining Repetitive Job Behavior
A library for detecting and measuring I/O performance variability.
Dataset of I/O Scalability Characterization for Serverless Applications
A framework for characterizing and mitigating I/O scalability bottlenecks in serverless applications.
SATORI: Bayesian Optimization Based Multi-Core Resource Partitioning Framework
A Bayesian optimization based resource partitioning framework for throughput-oriented workloads.
Failure Dataset for Two Generations of GPU-Accelerated Supercomputers
A failure dataset collected from two generations of GPU-accelerated supercomputers.
BLISS: Performance Auto-Tuning Framework for Parallel Applications
A performance auto-tuning framework for parallel applications.
QRAFT: Quantum Program Output Estimation Framework Using Reversibility
A quantum program output estimation framework based on reversibility.
GIFT: Coupon-Based I/O Bandwidth Allocation Framework
A coupon-based framework for allocating I/O bandwidth.
UREQA: Machine Learning Based Qubit Allocation for Quantum Programs
A machine learning based qubit allocation framework for quantum programs.
Wizard: Machine Learning Based Disk Failure Prediction Framework and Dataset
A disk failure prediction framework and disk failure dataset.
Veritas: Quantum Program Output Estimation Using Post-Processing of Noisy Runs
A quantum output estimation framework using post-processing of noisy runs.
Quantum Error and Benchmarking Dataset for IBM NISQ Machines
A dataset for quantum error characterization and benchmarking on IBM NISQ machines.
HPC Resource Consumption Dataset and Prediction Tool
A resource consumption dataset for multiple supercomputers with a prediction tool.
DISQ: Quantum Output State Classification Toolbox Using IBM OpenPulse
A toolbox for quantum output state classification using IBM OpenPulse.
CLITE: QoS Framework for Latency-Critical Microservices
A QoS framework for latency-critical microservices in data centers.
Power Consumption Traces of Real-World Applications on Production HPC Systems
Power consumption traces collected from real-world applications on production HPC systems.
SSD Vibration and Reliability Dataset
A dataset for studying SSD vibration and reliability behavior.
PERQ: Power Management Framework for HPC Systems
A power management framework for high-performance computing systems.