More Like this

1. Novel Network Services for Supporting Big Data Science Research

   Chung, J.; Donovan, S.; Bezerra, J.; Morgan, H.; Ibarra, J.; Clark, R.; Owen, H (October 2017, Gateways 2017)

   Poster Abstract: To interconnect research facilities across wide geographic areas, network operators deploy science networks, also referred to as Research and Education (R&E) networks. These networks allow experimenters to establish dedicated network connections between research facilities for transferring large amounts of data. Recently, R&E networks have started using Software-Defined Networking (SDN) and Software Defined Exchanges (SDX) for deploying these connections. AtlanticWave/SDX is a response to the growing demand to support end-to-end network services spanning multiple SDN domains. However, requesting these services is a challenging task for domain-expert scientists, because the interfaces of the R&E networks have been developed by network operators for network operators. In this paper, we propose interfaces that allow domain expert scientists to reserve resources of the scientific network using abstractions that focus on their data transfer needs for scientific workflow management. Recent trends in the networking field pursue better interfaces for requesting network services (e.g., intent-based networking). Although intents are sufficient for the needs of network operations, they are not abstract enough in most cases to be used by domain-expert scientists. This is an issue we are addressing in the AtlanticWave/SDX design: network operators and domain expert scientists will have their own interfaces focusing on their specific needs. 

   more » « less
2. Novel Network Services for Supporting Big Data Science Research

   Chung, J.; Donovan, S.; Bezerra, J.; Morgan, H.; Ibarra, J.; Clark, R.; Owen, H (October 2017, Gateways 2017)

   To interconnect research facilities across wide geographic areas, network operators deploy science networks, also referred to as Research and Education (R&E) networks. These networks allow experimenters to establish dedicated network connections between research facilities for transferring large amounts of data. Recently, R&E networks have started using Software-Defined Networking (SDN) and Software Defined Exchanges (SDX) for deploying these connections. AtlanticWave/SDX is a response to the growing demand to support end-to-end network services spanning multiple SDN domains. However, requesting these services is a challenging task for domain-expert scientists, because the interfaces of the R&E networks have been developed by network operators for network operators. In this paper, we propose interfaces that allow domain expert scientists to reserve resources of the scientific network using abstractions that focus on their data transfer needs for scientific workflow management. Recent trends in the networking field pursue better interfaces for requesting network services (e.g., intent-based networking). Although intents are sufficient for the needs of network operations, they are not abstract enough in most cases to be used by domain-expert scientists. This is an issue we are addressing in the AtlanticWave/SDX design: network operators and domain-expert scientists will have their own interfaces focusing on their specific needs. 

   more » « less
3. WfCommons: Datasets and Tools for Advancing Scientific Workflows

   https://doi.org/10.1145/3750720.3757301  

   Casanova, Henri; Coleman, Tainã; Ferreira_da_Silva, Rafael; Suter, Frédéric (September 2025, ACM)

   Scientific workflows have been and continue to be relied upon by thousands of Cyberinfrastructure practitioners for managing data analyses, simulations, and other computations in almost every scientific domain. To advance scientific workflow research and development, the WfCommons project aims to provide workflow datasets. These datasets allow researchers to obtain novel results by using the provided data to perform sound and informed evaluation of their algorithms, developers to transform the testing and evaluation of the runtime systems they develop, and educators to achieve new learning objectives by exposing students to compelling, real-world scenarios. The data items generated and curated by this work are called workflow instances: accurate descriptions of workflow specifications and executions on production systems, available in JSON using a well-defined schema. To make these workflow instances easily usable by Cyberinfrastructure practitioners, WfCommons also provides software tools for describing, constructing, analyzing, visualizing, and generating workflow instances and for generating and executing workflow benchmarks on production platforms. Some of the WfCommons tools come as Python packages with simple APIs, while others are usable directly in the browser. 

   more » « less
4. Enabling low-latency machine learning on radiation-hard FPGAs with hls4ml

   https://doi.org/10.1088/2632-2153/ae8218  

   Govorkova, Katya; García_Pardiñas, Julián; Lončar, Vladimir; Schmitt, Sebastian; Nguyen, Victoria; Pizzichemi, Marco; Martinazzoli, Loris; Anne_Smith, Eluned (July 2026, Machine Learning: Science and Technology)

   Abstract This paper presents an end-to-end demonstration of a viable, ultra-fast, radiation-hard machine learning (ML) application on FPGAs, which could be used in future high-energy physics experiments. We present a three-fold contribution, with the PicoCal calorimeter, planned for the Large Hadron Collider beauty Upgrade II experiment, used as a test case. First, we develop a lightweight autoencoder to compress a 32-sample timing readout, representative of that of the PicoCal, into a two-dimensional latent space. Second, we introduce a systematic, hardware-aware quantization strategy and show that the model can be reduced to 10-bit weights with minimal performance loss. Third, as a barrier to the adoption of on-detector ML is the lack of support for radiation-hard FPGAs in the High-Energy Physics community’s standard ML synthesis tool,hls4ml, we develop a new backend for this library. This new back-end enables the automatic translation of ML models into high-level synthesis projects for the Microchip PolarFire family of FPGAs, one of the few commercially available and radiation hard FPGAs. We present the synthesis of the autoencoder on a target PolarFire FPGA, which indicates that a latency of 25 ns can be achieved. We show that the resources utilized are low enough that the model can be placed within the inherently protected logic of the FPGA. Our extension tohls4mlis a significant contribution, paving the way for broader adoption of ML on FPGAs in high-radiation environments. 

   more » « less
5. Researcher reasoning meets computational capacity: Machine learning for social science

   https://doi.org/10.1016/j.ssresearch.2022.102807  

   Lundberg, Ian; Brand, Jennie E.; Jeon, Nanum (November 2022, Social science research)

   Computational power and big data have created new opportunities to explore and understand the social world. A special synergy is possible when social scientists combine human attention to certain aspects of the problem with the power of algorithms to automate other aspects of the problem. We review selected exemplary applications where machine learning amplifies researcher coding, summarizes complex data, relaxes statistical assumptions, and targets researcher attention to further social science research. We aim to reduce perceived barriers to machine learning by summarizing several fundamental building blocks and their grounding in classical statistics. We present a few guiding principles and promising approaches where we see particular potential for machine learning to transform social science inquiry. We conclude that machine learning tools are increasingly accessible, worthy of attention, and ready to yield new discoveries for social research. 

   more » « less