News

As we approach the end of November, check out all the interesting things that have been happening with #MaterialsDataMonth on Twitter and LinkedIn!

Globus and Glabs had a great week at SC22 in Dallas! You can check out all of our talks, papers, and panels here.

Ian accepted his ACM Ken Kennedy Award and gave his Ken Kennedy Lecture at SC22!

Congrats to Greg, Valerie, Logan, Ian, and collaborators for winning the Gordon Bell Special Price for COVID-19 Research at SC22! Check out the prize-winning paper: “GenSLMs: Genome-Scale Language Models Reveal SARS-CoV-2 Evolutionary Dynamics”! Several other GLabs projects were used in this work, including Parsl, Colmena, and ProxyStore!

Logan is now staffing the Informatics Help Desk on Fridays! Stop by for any of your informatics needs or check out the website for upcoming workshops and tutorials.

Parsl was selected by the Chan-Zuckerberg Initiative for funding to support essential open source software for science!

Congrats to Ben, Ryan, Ari, KJ, Kyle, and Ian for winning HPCwire’s Best Use of High Performance Data Analytics & Artificial Intelligence Award for 2022. Check out the relevant paper on FAIR principles for AI here and the related blog post here!

Max moderated at panel on High-level IRs for a C/C++ Optimizing Compiler at the 2022 LLVM Developers’ Meeting.

Tak’s paper “AICCA: AI-driven Cloud Classification Atlas” was accepted in MDPI, Remote Sensing! A preprint is available on arXiv.

Eamon’s paper “Instruments, Agents, and Artificial Intelligence: Novel Epistemic Categories of Reliability” was accepted to Synthese for a special issue on “Philosophy of Science in Light of Artificial Intelligence”.

Andre Bauer has joined us from the University of Wurzburg as our newest PostDoc! Andre did his PhD on self-adaptive systems and time series forecasting. You can check out Andre’s past work on his Google Scholar page!

Tak had 2 papers accepted as posters at NeurIPS!

Matt presented his paper “FLoX: Federated Learning with FaaS at the Edge” at eScience 2022.

Matt presented a poster of his paper “Exploring Tradeoffs in Federated Learning on Serverless Computing Architectures” at eScience 2022.

Kyle presented “Globus: A Platform for Global Data Management” at the NSDF meeting.

Ian presented at ESnet6 launch event on “How ESnet6 enables new smart instruments for21st Century science”.

Ian delivered the keynote at eResearch Australia on the new Patterns paper!

Eamon was named a “Rising Star of Data Science” by the Data Science Institute at the University of Chicago.

OMG! Daniel presented “OMGMO: Original Multi-modal Dataset of Genetically Modified Organisms in African Agriculture” this weekend at the International Conference on Social Informatics in Glasgow.

Eamon was interviewed by Nature regarding the role of AI in the peer review process! Check out the full article here.

Nathaniel has been selected to serve on 2 TPCs for PERCOM’s WiP 2023 session and DCOSS-IoT 2023.

Ian is the 2022 recipient of the prestigious IEEE-ACM Ken Kennedy Award!!

We have six new PhD students join the group: Daniel Grzenda, Alok Kamatar, Arhan Khan, Mansi Sakarvadia, Michael Tynes, and Wenyi Wang!

Eamon will be presenting his paper, “Deep Learning Opacity in Scientific Discovery”, at the 120th annual meeting of the Central Division of the American Philosophical Association.

Logan will be presenting at the Computing in Engineering Forum at UW-Madison regarding our work on designing batteries.

Logan is presenting at UW-Madison’s Computing in Engineering Forum

Ryan, Raf, and Ian contributed to a paper “Fixed-target serial crystallography at the Structural Biology Center”

Matt’s paper on “Rural AI: Serverless-Powered Federated Learning for Remote Applications” was accepted to IEEE Internet Computing

Ian is giving a keynote presentation at the Harnessing Data for Materials Symposium.

Eamon is presenting “Optimizing AI for the Human Loop” at the Harnessing Data for Materials Symposium.

The automation/gladier team has shared two new preprints: on Globus automation services and on patterns for linking experiments and HPC

Jim and Ian (in collaboration with Justin Wozniak) had a paper accepted to the ReWords workshop at eScience.

Ian, Nathaniel and Valerie will be attending the DOE’s AI for Science, Energy and Security workshop in Bowie, MD.

Tak and his colleagues win the Unexpected Discovery award (one of the top 3 awards!) at the Frontier Development Lab 2022 live showcase for their work on environmental remediation. Congrats Tak 🎉 !

Nathaniel Hudson joined our team as a post-doc after finishing his PhD at the University of Kentucky!

Yuanjian’s paper on an optimized framework for error-bounded lossy compression is now published in IEEE Transactions on Parallel and Distributed Systems.

Ian co-authored a paper on an approach to apply FAIR principles in a continuous and ubiquitous manner in the Harvard Data Science Review.

Max presented to the Circuit IR Compilers and Tools (CIRCT) Open Development Meeting on his work to compile neural networks to FPGAs.

Ben and Ian co-authored a paper on needs for analyzing microscopy datests with our friends at UW-Madison.

The Globus Labs team part of two new NSF CSSI grants:

• Garden - A new project which will make ML/AI models accessible to more scientists
• Continuing support for Parsl, an open source library for parallel Python on HPC

Ian and colleagues published a perspective in Science on how slow science is to build open research infrastructure.

Ben Blaiszik gave the keynote at SciPy 2022 on the impact of open Python on scientific research.

Eamon’s paper on whether the opacity of deep learning models inhibits scientific discovery (it doesn’t™️) was accepted to Philosophy of Science

Globus Labs will be hosting Parsl and FuncXFest this year as a virtual workshop. We hope to see you there!

Two of our incoming team members, Mansi and Mike, were featured in articles about their CSGF fellowships. Congratulations to them!

An interview of Ian and Rosa Badia on how AI and simulation will combine to solve global challenges is on IEEE Computer.

Kyle discussed serverless scientific workflows at the 7th Workshop on Serverless Computing

Logan’s interview on the ExaLearn project is now published as a podcast

Ian presented how Globus is enabling global research infrastructure to National Science Data Fabric. Video is on our YouTube

Ben Blaiszik will deliver the keynote at the SciPy 2022 on software and services to fully realize the impact of AI, Machine Learning, automation, and community engagement in science.

Ben Blaiszik spoke at the FAIR for AI Workshop on the relationship between data and artificial intelligence.

Kyle spoke at the NREL Connections Seminar on Scalable Scientific Computing with funcX and Parsl

Ian was one of the distinguished speakers at National Science Data Fabric, speaking on “How Globus Services Enable Scientific Discovery” (recording here).

A preprint of our first paper on ScholarBERT “ScholarBERT: Bigger is not always better” by Hong, Aswathy and Greg et al is now published on arXiv. The pretrained ScholarBERT models are available on HuggingFace Hub!

Matt’s paper on Federated Learning, titled “Adaptive Edge-Cloud Environments for Rural AI” has been accepted to IEEE SCC 2022!

Dr. Tyler J. Skluzacek successfully defended his dissertation: “Automated metadata extraction can make data swamps more navigable” to his committee of Kyle Chard, Ian Foster, Raul Castro-Fernandez, and Michael Franklin! Globus Labs member Logan Ward raved about Skluzacek’s defense performance, “Nicely done, Tyler! I was very impressed by the clarity and charisma of your presentation”. Committee member Michael Franklin raved, “He brought muffins. Pass!” It was truly an unmissable, five-star spectacle: ⭑⭑⭑⭑⭑.

Ben Blaiszik and Ari Scourtas took part in the CHiMaD workshop on data management, standards, reproducibility, and ML-ready materials data.

Ian, Kyle, and others from the Globus team present the state of multiple Globus Labs research projects at Globus World ‘22.

Globus World ‘22 takes place on May 10 and 11 in both 3D (in person in Chicago) and 2D (virtual attendance). Ian will co-present the keynote at 9:00AM (Central) on May 10 and Kyle will co-present “Shaping the Future: To Globus Compute and Beyond!” at 4:00PM.

Student researchers Jim and Nathan along with Marcus, Ben and Ian co-authored “3D Convolutional Neural Networks for Dendrite Segmentation Using Fine-Tuning and Hyperparameter Optimization.” Pre-print is available on Arxiv with journal publication to follow.

Ian co-authored "Efficient high-dimensional variational data assimilation with machine-learned reduced-order models" which appears in [_Geoscientific Model Development_](https://gmd.copernicus.org/articles/15/3433/2022/).

Kyle gave a Presentation on Parsl at Software & Computing Round Table.

Tyler Skluzacek will start a new position as a Research Scientist at Oak Ridge National Laboratory this summer. Congrats Tyler! 🎊

Tyler Skluzacek will defend his thesis “Automated Metadata Extraction Can Make Data Swamps More Navigable” on May 20th. 🎉

Globus and Globus Labs will be a recipient of the National Science Foundation’s CONECT award under the [Advanced Cyberinfrastructure Coordination Ecosystem Services and Support (ACCESS) program].

Zhi Hong, J. Gregory Pauloski, Aswathy Ajith, Eamon Duede, Kyle Chard and Ian Foster will present “Accelerating Computational Social Science with ScholarBERT” at the 8th International Conference on Computational Social Science (IC2S2) at the University of Chicago, July 19-22.

Eamon Duede will present “Distance Matters for Science and Scholarship” at the 8th International Conference on Computational Social Science (IC2S2) at the University of Chicago, July 19-22.

Lab members organized and ran the “Fueling Materials Discovery with Machine Learning and FAIR Data Approaches” 2-hour training workshop at the 3M Tech Forum. This included presentations and research content from Logan Ward, Marcus Schwarting, Ari Scourtas, KJ Schmidt, Ben Blaiszik, Ian Foster, Ben Galewsky, Kyle Chard, Ryan Chard, and Nick Saint.

Logan Ward presented on “Resource-Efficient Material Design Campaigns with Artificial Intelligence” at Arizona State University as part of their AI + Materials Spring 2022 Seminars series.

Ben Blaiszik hosted the first of many YouTube Lives that will center around conversations with researchers focused on data, AI/ML, and #OpenScience! This stream was with Assistant Scientist Logan Ward on “A Starter Code Template for Scientific AI Projects.” You can access the stream recording on YouTube and the associated notebook on GitHub. Follow Ben for announcements of future sessions!

Ian’s 2011 paper “The History of the Grid,” originally in “High Performance Computing: From Grids and Clouds to Exascale”, is now freely available on arXiv!

KJ and Ben gave a seminar at University of Utah - Department of Materials Science & Engineering–“Introduction to Scientific Data Infrastructure”!

Check out our new preprint addressing FAIR (Findable, Accessible, Interoperable, and Reusable) data in materials science on arXiv!

Tyler presented “Automated metadata extraction to make data swamps more navigable” at Oak Ridge National Laboratory.

Kyle Chard with a volunteer team of researchers led by Rafael Ferreira da Silva of the U.S. Department of Energy’s Oak Ridge National Laboratory (ORNL) launched the Workflows Community Initiative (WCI).

Read the full article at: https://www.hpcwire.com/off-the-wire/ornl-and-uchicago-lead-the-launch-of-unique-workflows-community-platform/

Argonne Leadership Computing Facility features the work of Marcus Schwarting, Noah Paulson, Ben Blaiszik, Ian Foster. “Argonne and 3M use machine learning to enhance material manufacturing process” on 3M work.
Read the full article at:

https://www.alcf.anl.gov/news/argonne-and-3m-use-machine-learning-enhance-material-manufacturing-process

Max presented his work on the Torch-MLIR community meeting. Read the agenda here: https://discourse.llvm.org/t/torch-mlir-community-meeting-2022-04-04/60978 and the full PR for the work here: https://github.com/llvm/torch-mlir/pull/635

Our students will migrate to silicon valley for the summer. Hong will be interning at Amazon as an Applied Scientist intern with the Personalization team. Aswathy will be at Microsoft as a data science intern with the devices team; and Yuanjian will be at Google as a software engineer with the ads team

Kyle Chard’s co-authored paper “SCANNS: Towards Scalable and Concurrent Data Indexing and Searching in High-End Computing System” was accepted at CCGrid 2022.

Logan Ward presented (virtually) at TMS 2022 on software approaches for making laboratories fully autonomous. You can see his talk on YouTube

Max’s paper “Ultrafast Focus Detection for Automated Microscopy” accepted for publication to ICCS 2022.

Roselyne Tchoua named Sinai Urban Health Institute Equity Research Fellow. Congratulations Rosalyne!

Tyler’s paper “Models and Metrics for Mining Meaningful Metadata” accepted for publication to ICCS 2022.

Ben, Logan, Ari, and KJ led a training in the Argonne Leading Computing Facility (ALCF) AI for Science Training series, in collaboration with Eliu Huerta and Venkat Vishwanath of Argonne National Laboratory. Trainees learned how to leverage modern computing environments and advanced scientific data infrastructure (DLHub, funcX and ThetaGPU) to create FAIR (Findable, Accessible, Interoperable, and Reusable) AI models and datasets. Video is available here.

Zhi Hong presented a poster titled “Materials Science language models and analysis” at the Materials Research Data Alliance (MaRDA) Annual Meeting 2022.

Globus World will be held May 10-12, 2022 in Chicago. In this year, we are back to in-person sytle meeting. Go Globus!

Eamon presented his work at Social Science Foo Camp!

Logan and Ben serve as panelists at the Materials Research Data Alliances (MaRDA) Annual Meeting 2022.

Ben presented “A FAIR Approach Towards Fully Realizing the Impact of AI and Machine Learning in Materials Science” and was on a panel for the CECAM Machine actionable data for chemical sciences: Bridging experiments, simulations, and machine learning for spectral data (MADICES) conference.

Yadu presented Parsl to a group at the Princeton Plasma Physics Laboratory.

Eamon presented “Being Together in Place as a Catalyst for Scientific Advance” at the Seminar on Economic Geography.

Raf’s paper “Materials Engineering of Violin Soundboards by Stradivari and Guarneri” made the cover of Angewandte Chemie

Hong, Ari, Ben presented at Chimad.

Eamon’s paper “How status of research papers affects the way they are read and cited” was published in Research Policy.

Ben will be giving at Keynote at CECAM on “A FAIR Approach Towards Fully Realizing the Impact of AI and Machine Learning in Materials Science”.

Ben gave the NIST Office of Data Informatics Monthly Seminar - “A FAIR Approach Towards Fully Realizing the Impact of AI and Machine Learning in Materials Science”

ANL multi-division collaboration article on PFAS toxicity “Uncertainty-Informed Deep Transfer Learning of Perfluoroalkyl and Polyfluoroalkyl Substance Toxicity” has been published in The Journal of Chemical Information and Modeling.

Tyler’s study on how reserachers find and use open data has been approved by an IRB! Contact him if you’ve got a perspective to share on how data could be made more discoverable.

Eamon Duede was awarded an NSF funded traineeship at the Pritzker School of Molecular Engineering for AI-Enabled Molecular Engineering of Materials and Systems for Sustainability. Congratulations!

We are glad to be part of the recent publication on common drivers and databases for quantum chemistry lead by Lori Burns.

The multi-lab effort for COVID drug screening was published in The Journal of Chemical Information and Modeling. Congratulations, team!

Ian is taking part in the NNSA’s committee on post-exascale computing. We are already planning for the next era of computing!

Ryan Wong won the 2nd place with Delving Into the Abyss: A Distributed Decompression System for Indexing Compressed Repositories. Kevin Gao won the 3rd place with Regulating Traffic in a Crowded Cache: Overcoming the Container Explosion Problem. Congratulations!

Ganesh presented “From Atomistic to Coarse Grained : Active Learning Strategies for Gaussian Approximation Potential and Deep Kernel Learning” at the ML-IP 2021 Psi-k tutorial workshop .

Tyler was named a UChicago Data Science Institute Rising Star in their Autumn 2021 cohort. Congratulations Tyler!

In collaboration with Cardiff University, Kyle’s paper “Greedy Nominator Heuristic (GNH): Virtual Function Placement on Fog Resources” was accepted to the Concurrency and Computation: Practice and Experience journal!

Check out tutorials, papers, and presentations by Globus Labs members at SC21! A full schedule of Globus and Globus Labs events at SC21 is available at globus.org/events/sc21.

A new paper with Ewa Deelman won best paper at FGCS.

Ben presented “Advanced Robotics and Software Defined Automation to Speed Materials Discovery” at the AI/ML for Particle Accelerator, X-Ray Beamlines and Electron Microscopy workshop.

Ethan Schondorf joins Globus Labs!

Kyle and Raul’s A Community Resource for Innovation in Polymer Technology (CRIPT) (in collaboration with Bradley Olsen at MIT) won funding through the NSF Convergence Accelerator.

Some of Logan’s thesis work has finally been published: High-throughput crystal structure solution using prototypes. Congratulations Logan!

Ben and Raf and Ryan are contributing to ALCF’s AI for Science Training Series with a tutorial on Globus. Time and date TBD.

Logan and Ryan are contributing to ALCF’s AI for Science Training Series. Logan is presenting on 11/04/2021, 3:00PM and Ryan on 02/17/2022, 3:00PM.

Ian Foster wrote a new Globus blog post titled “Extra-fast Export: Globus Exceeds on Exascale Day by Accelerating Big-file Transfers”. Check out the post to learn how Globus transferred data from NERSC to ALCF at 30 gigabits per second!

Tak is presenting Cloud Clustering Applied to MODIS Calibrated Radiances Over 2003 via Rotationally Invariant Autoencoder at AGU21 Wednesday, 15 December 2021.

Our very own Logan Ward has begun his transition from research scientist to becoming the next Carl Sagan… Just kidding! Logan’s recordings from his Applied AI for Materials Science taught at UChicago during Winter 2021 are now available.

Parsl & funcX Fest 2021 was a great success this year. Thanks for all that contributed to its smooth running and all those that attended.

Greg’s KAISA: An Adaptive Second-Order Optimizer Framework for Deep Neural Networks has been accepted into SC this year. Congratulations Greg!

Yuanjian’s Understanding Effectiveness of Multi-error-bounded Lossy Compression for Preserving Ranges of Interest in Scientific Analysis has been accepted into the 7th International Workshop on Data Analysis and Reduction for Big Scientific Data. Congratulations Yuanjian!

Tong Shu’s In-situ Workflow Auto-tuning via Combining Performance Models of Component Applications has been accepted into SC this year. Congratulations Tong!

Matt’s Embeddings are All You Need: Transfer Learning in Convolutional Neural Networks using Word Embeddings is at SRC at SC this year.

Many papers this year at 16th Workshop on Workflows including: ExaWorks: Workflows for Exascale by Al-Saadi et al, Not All Tasks Are Created Equal: Adaptive Resource Allocation for Heterogeneous Tasks in Dynamic Workflows by Phung et al, A Community Roadmap for Scientific Workflows Research and Development by Ferreira da Silva et al.

Several students posters this year at SC21 including: Regulating Traffic in a Crowded Cache: Overcoming the Container Explosion Problem by Kevin Gao, Delving Into the Abyss: A Distributed Decompression System for Indexing Compressed Repositories by Ryan Wong, Mitigating the Metadata Mess: Autonomous Metadata Extraction Pipelines for Large-Scale Data Repositories by Matthew Chen and Erica Hsu.

Ian will be delivering the keynote at ScalA21: 12th Scalable Algorithms for Large-Scale Systems on November 19, 2021 (keynote abstraction here).

Ben and Ian’s work on the Delta variant and Vaccine efficacy is available as a preprint

Big group effort on Exascale Machine Learning has been published in The International Journal of High Performance Computing Applications.

Hong’s “Challenges and Advances in Information Extraction from Scientific Literature” has been accepted into The Journal of The Minerals.

Tekin’s “High-Performance Ptychographic Reconstruction with Federated Facilities” has been accepted into SMC2021.

Globus Labs welcomes new postdoc Valerie Hayot! Valerie recently completed her Ph.D. in Software Engineering at Concordia University (Canada).

Globus Labs congratulates the following students on their poster acceptances to SC21: Kevin Gao, Matt Baughman, Matthew Chen + Erica Hsu (joint), and Ryan Wong.

Kyle Chard presented “Serverless Supercomputing: High Performance Function as a Service” in the CloudHPC symposium.

Ian Foster speaks on panel entitled “Exploring the growing synergy between cloud and high performance computing” at the CloudHPC symposium.

Congrats to students in the 2021 CDAC Data and Computing Summer Lab for completing their 2021 lab experiences! You can see the students and their projects here.

Matt Baughman (M.S.) and Tyler Skluzacek (Ph.D. candidacy) successfully passed their respective exams. Congrats, Matt and Tyler!

The BigDataX REU has finished! Congrats to Jamie, Kevin, and Matthew on a successful summer of research.

“Analysis of In-Situ X-Ray Tomography Datasets of Dendritic Solidification Using 2D and 3D Machine Learning Algorithms” Tiberiu Stan, Nathan Pruyne, Jim James, Marcus Schwarting, Jiwon Yeom, Ben Blaiszik, Ian Foster, Peter Voorhees

Hong paper “Models and Processes to Extract Drug-like Molecules from Natural Language Text” was accepted in Frontiers Molecular Biosciences.

Ian’s paper “Autonomous experimentation systems for materials development: a community perspective” is published in Matter.

Matt (Enhancing Automated FaaS with Cost-aware Provisioning of Cloud Resources), Tak’s (Cloud Clustering Over January 2003 via Scalable Rotationally Invariant Autoencoder), Hong’s (Gold Panning: Automatic Extraction of Scientific Information from Publications), Max’s (Ultrafast Focus Detection for Automated Microscopy), and the funcX teams’ (Federated Function as a Service for eScience) posters were accepted to eScience 2021.

Tak’s paper “Data-driven Cloud Clustering via a Rotationally Invariant Autoencoder” (DOI: 10.1109/TGRS.2021.3098008) is accepted for publication at IEEE Transactions on Geoscience and Remote Sensing!

Logan Ward’s paper “Graph-Based Approaches for Predicting Solvation Energy in Multiple Solvents: Open Datasets and Machine Learning Models” is accepted for publication at The Journal of Physical Chemistry A!

Ben and his co-investigators have received $297,000 from DOE’s High Performance Computing Research at U.S. National Laboratories program for their project “Next Generation Nonwovens Manufacturing Based on Model-Driven Simulation Machine Learning Approach”. They will use a combination of HPC based CFD simulations and a machine learning to minimize energy consumption of melt blown fiber manufacturing processes. See the official announcement here.

Kyle speaks at Internet 2 Community Voices Series on “Science Gateways: Sustaining Integral Solutions for Accelerating Science”. Recording to follow soon!

Globus Labs has been awarded OAC Core Research funding with Illinois Tech for “Enabling Extremely Fine-grained Parallelism on Modern Many-core Architectures.”

Tyler Skluzacek presented Globus Labs paper entitled “A serverless framework for bulk metadata extraction” at HPDC 2021.

We are excited to announce two exciting tutorials for the ExaWorks project and for Globus at SC21! More details to follow soon.

Congratulations to Tong Shu and the entire CODAR Team on the acceptance of a paper to be presented at SC21.

Greg had his paper titled “KAISA: An Adaptive Second-order Optimizer Framework for Deep Neural Networks” accepted at SC21! Stay tuned for more information.

Globus Labs has had two papers accepted at eScience 2021—“An Empirical Study of Package Dependencies and Lifetimes in Binder Python Containers” and “Extreme Scale Survey Simulation with Python Workflows” in collaboration with DESC! More details to follow.

Globus Labs and the UChicago Center for Data and Computing would like to welcome the 55 high school, undergraduate, and graduate students joining us this summer for the CDAC Summer Lab! Check out the program, students, and project profiles on the CDAC Summer Lab website here.

Check out Ben’s new paper on additive manufacturing modeling in Integrating Materials and Manufacturing Integration!

Globus Labs welcomes new Ph.D. student Aswathy Ajith! She will be working on information extraction systems for scientific data using natural language processing techniques.

On June 22nd, Tyler Skluzacek will present the Globus Labs paper entitled “A serverless framework for bulk metadata extraction” at HPDC 2021.

After some fiddling around, Raf’s collaboration paper on the structure of violins – “Materials engineering of violin soundboards by Stradivari and Guarneri” – has been accepted by Angewandte Chemie, one of the top three chemistry journals in the world!

Globus Labs welcomes our BigDataX REU students. BigDataX is co-hosted by the UChicago and IIT.

Max’s, Ryan’s, Ben’s, and Ian’s co-authored paper “Accelerated, Scalable and Reproducible AI-driven Gravitational Wave detection” is accepted for publication at Nature Astronomy!

Raf Vescovi and Ryan Chard present at the virtual 2021 APS/CNM Joint Virtual Users Meeting.

Logan Ward gave a tutorial talk on applying AI to molecular simulations to the MICCOM Center.

Logan Ward is presenting a tutorial on optimal experimental design techniques as part of Argonne’s AI for Science Tutorial series.

Globus Labs alumni Tim Shaffer and Rohan Kumar present papers at IPDPS 2021. Tim presented “Lightweight Function Monitors for Fine-grained Management in Large Scale Python applications” in the main IPDPS track, and Rohan presented his paper “Coding the Computing Continuum: Fluid Function Execution in Heterogeneous Computing Environments” at the HCW21 workshop.

Ian and Kyle hosted a seminar in Serverless Computing as part of the Dagstuhl Seminar series.

Check out this Argonne press on “Ganesh’s PRL”. The title “Lasers, levitation and machine learning make better heat-resistant materials” speaks by itself. The collaboration between Chris Benmore from the APS and our Data Science team is making ripples on material science. Check the full article “here”.

Exciting news. Kyle got a new funding in collaboration with IIT that will support a developer and another student at University of Chicago. The goal is to achieve A High-Performance Storage Infrastructure for Activity and Log Workloads on Frameworks. As always we have an amazing name for the project, ChronoLog.

Raf presented the first operational version of Gladier: The Globus architecture for data-intensive experimental research at the “Autonomous Discovery in Science and Engineering” organized by CAMERA.

Ian and Kyle bring globus to NZ. You can check the webinar at “Nesi - Globus Update for professionals”

Ari will be giving a short 5-min presentation on DLHub and Foundry at the “Research Data Alliance’s 17th Plenary”, during the Defining FAIR for Machine Learning session this Wednesday (4/21) at 15:00 UTC (10am CT), with a recorded repeat on Thursday (4/22) at 23:00 UTC (6pm CT).

Check out all the new features released in the newest version of funcX! funcX is our FaaS framework specifically designed for high performance function serving.

The “Computing Comunity Consortium” highlights Ian’s white paper on the National Discovery Cloud. The article highlights the three pillars of computational fabric.

You can check the full paper “here”

Raf presented the first operational version of Gladier: The Globus architecture for data-intensive experimental research at the “Autonomous Discovery in Science and Engineering” organized by CAMERA.

Ganesh and the Argonne team had a paper titled “Experimentally Driven Automated Machine-Learned Interatomic Potential for a Refractory Oxide” published in the PRL Journal! This paper was also highlighted as an “Editor’s Suggestion” in the journal.

Logan and Ganesh, in collaboration with others, had a new paper accepted to the Ninth International Conference on Learning Representations! You can check out a preprint of the paper, titled “Benchmarking Deep Graph Generative Models for Optimizing New Drug Molecules for COVID-19”, on arXiv.

In collaboration with the Center for Data and Computing, we are hosting the Chicago Data Discovery Virtual Summit on April 22 and 23! Registration is free and we look forward to seeing you there.

A deployment of the Modern Research Data Portal (MRDP), a high-performance networking resource developed by engineers at the U.S. Department of Energy’s Energy Sciences Network (ESnet) in collaboration with computer scientists at Argonne National Laboratory and the University of Chicago, is now being used at NERSC for the Dark Energy Science Collaboration (DESC)!

Check out all the new features released in the newest version of funcX! funcX is our FaaS framework specifically designed for high performance function serving.

Check out a new preprint from a Globus Labs collaboration: High Throughput Virtual Screening and Validation of a SARS-CoV-2 Main Protease Non-Covalent Inhibitor!

Rohan Kumar’s “Coding the Continuum: Fluid Function Execution in Heterogeneous Computing Environments” has been accepted to the Heterogeneity in Computing Workshop at IPDPS 2021!

Ian Foster, Kyle Chard, Yadu Babuji, and Zhuozhao Li are chairing the Workshop on High Performance Serverless Computing which will be held virtually in conjunction with HPDC2021

Ben Blaiszik and Ian Foster helped organize and plan the first annual meeting of the Materials Research Data Alliance (MaRDA). Topics included FAIR materials data, identification of state of the art data infrastructure, gap analyses, panels and presentations from leading experts in materials data, and more. The full program can be found here. Recorded videos will be available soon.

Ganesh Sivaraman’s paper, entitled “Experimentally driven automated machine-learned interatomic potential for a refractory oxide”, has been accepted for publication in the journal Physical Review Letters.

The DOE Secretary’s Honor Award names multiple Globus Labs members for numerous recent computing accomplishments surrounding COVID-19.

Aristana Scourtas has joined Globus Labs as a researcher. Ari is working on applied machine learning in chemistry, catalysis, and materials science and building data infrastructure to make it simple to discover and consume ML-ready data and models as part of Globus Labs. She aims to take on new projects applying machine learning and data science to environmental problems.

Ian was on the team that wrote “Fast and accurate learned multiresolution dynamical downscaling for precipitation” (Jiali Wang, et al). Check out the preprint here!

Course materials for Logan’s new class, “Applied Artificial Intelligence for Materials Science and Engineering,” at University of Chicago, have been posted. Check them out here!

KJ Schmidtjoins from Northwestern, and will be working on applied machine learning applied to problems in soft materials, atomic force microscopy, and more with industrial partners and general data infrastructure projects as part of Globus Labs.

The Exaworks project has been funded for another year. Congratulations, exaworks team!

Ian Foster and Anna Woodard are part of an interdisciplinary team which was awarded a 2020 Cancer Spotlight Grant of $100,000 by the UChicago Medicine Comprehensive Cancer Center Cancer Research Foundation’s Women’s Board. Their project aims to use deep learning on magnetic resonance imaging and genomic data to predict which women are at higher risk of breast cancer and would benefit from more frequent screening, and which could safely reduce their screening frequency. Such a system could lead to more cancers being caught early, when they are curable.

Logan Ward and his co-investigators Meimei Li and Abdellatif Yacout have received $200,000 from Argonne National Lab’s Energy and Global Security directorate for their project “Accelerating Discovery and Development of New Nuclear Materials.” They will build an AI-based tool to characterize and forecast the radiation tolerance of new materials.

Yongyan Rao joins Globus Labs as a Software Engineer.

Parsl and Swift/T were featured in an article by Rob Farber about ExaWorks. Check out the article here: Workflow technologies impact SC20 Gordon Bell Covid-19 award winner and two of the three finalists.

Visiting Student Tim Shaffer had his paper, “Lightweight Function Monitors for Fine-Grained Management in Large Scale Python Applications”, accepted to be presented and published in the proceedings of IPDPS ‘21.

Tak presented his poster A Data-Driven Cloud Classification Framework Based On A Rotation-Invariant Autoencoder at the 2nd NOAA Workshop on Leveraging AI in Environmental Sciences!

Tak successfully defended his Masters Thesis: “A Data-driven Novel Cloud Classification Framework based on An Efficient Cloud Representation via Rotational and Non-Rotational Invariant Autoencoder.” Congrats Tak!

Parsl and Funcx were featured by Dennis Gannon in the eScience Group’s blog. Check out the article here: “A Look at Parsl and FuncX: Two Excellent Parallel Scripting Tools for Clouds and Supercomputers.” along with some great examples using these tools!

Ian Foster and the CODAR Group–Center for Online Data Analysis and Reduction–had their paper, Jieyang Chen et al., Accelerating Multigrid-based Hierarchical Scientific Data Refactoring on GPUs, IPDPS 2021., accepted to be presented and published in the proceedings of IPDPS ‘21.

Tak presented his poster A Data-Driven Cloud Classification Framework Based On A Rotation-Invariant Autoencoder at AGU Fall Meeting 2020!

Tyler Skluzacek’s app, NightWare, won the FDA approval! The news is spotlighted by UChicago CS Home Page and NPR.org!

Aarthi is presenting a poster in the student SC20 poster competition. Good luck, Aarthi!

Kyle will be presenting on a panel on language support for parallel and distributed computing at HILT 2020.

Ryan is presenting his work on Gladier this Friday at the SC20 Workshop on Urgent HPC.

Globus labs has 2 papers in the XLOOP Workshop at SC20: Max is presenting on controlling manufacutring with AI and Tekin is presenting on [streaming tomographic reconstruction]

Greg will be presenting his paper on distributed training of CNNs at SC20 next Thursday!

Ian, Valeriu Codreanu (SURFsara), and Zhao Zhang (TACC) are running the 5th Deep Learning on Supercomputers Workshop at SC20.

The paper ‘Towards Online Steering of Flame Spray Pyrolysis Nanoparticle Synthesis’ (preprint), led by Maksim Levental, was accepted at the XLOOP 2020 Workshop at SC20. Congratulations, Max!

Takuya Kurihana’s poster, A data-driven cloud classification framework based on a rotationally invariant autoencoder was accepted to be presented at the AGU fall meeting. Congratulations, Tak!

Ian Foster was interviewed by Wired magazine about how machine learning is being used to speed up battery development: AI Is Throwing Battery Development Into Overdrive.

Ian Foster was featured in the Material Data Matters podcast about how materials are related to big data. Check it out here!

ParslFest 2020 was a huge success! This second Parsl Community Meeting was held October 6-7. The two day meeting included presentations and discussions on how Parsl is being used, new features, and tutorials. If you missed out the fun, don’t worry– slides and videos for all presentations have been posted!

Check out the fantastic research being done by CDAC summer lab students!

Globus moves one exabyte of data! Read more here

Ari Scourtas joins Globus Labs as a Software Engineer.

Oak Ridge Leadership Computing Facility (OLCF), a multi-institutional and multi-disciplinary team led by Eliu Huertais, nominated for HPCWire’s Best Use of High Performance Data Analytics & Artificial Intelligence.

Greg Pauloski, Marcus Schwarting, Priyanka V. Setty, and Yuanjian Lui join Globus Labs as PhD students. Greg earned his BS in Computer Science from the University of Texas at Austin. He is interested in systems that enable machine learning research in high-performance environments. Marcus has worked at Argonne National Laboratory since 2018. He earned his BS’s in Mathematics and Chemical Engineering from the University of Louisville and his MS in Computer Science from the Georgia Institute of Technology. He is interested in machine learning applied to materials science, chemistry, and fluid dynamics. Priyanka has worked at Argonne National Lab for a year after earning her BE in Biotechnology from P.E.S.I.T, Bangalore, India and MS in Biostatistics from the University of Pittsburgh. She is interested in AI for answering biological problems. Yuanjian earned his BS in Computer Science from Zhejiang University. He is interested in distributed systems and loves filmmaking and video production. He is not super good at English, but if you have enough patience, you are likely to make friends with him.

DLHUb and Ben Blaiszik featured in DOE and NCSA Supercomputing Speed Studies into Black Hole Origins.

An Argonne LDRD was awarded to Dr. Kyle Chard to pursue work on integrating Parsl and Balsam to address big data challenges with x-ray diffraction technologies at the Advanced Photon Source.

An Argonne LDRD was awarded to Dr. Ryan Chard to work on real-time IoT fusion to drive machine learning for experiment steering and on-line optimization.

The Data Stations project begins with an aim to eliminate hurdles of big data exchange. Dr. Ian Foster states, “The Data Station is a radically new approach that is needed to change how people and organizations thing about, access, and use data”. This work will be funded by a $1 million grant from the National Science Foundation as part of the Convergence Accelerator program.

On the 17th of September, Dr. Ian Foster presented a talk, “‘AI for Science’ program at Argonne National Laboratory”, at the 2nd NOAA Workshop on Leveraging AI in Environmental Sciences.

funcX and Ryan Chard featured in News and Articles on Science and Technology Phys.org about using supercomputers to conduct real-time analysis of COVID-19 proteins.

The second ParslFest community meeting will be held as a (free) virtual meeting over two sessions (Oct 6 morning and Oct 7 afternoon). The meeting will include presentations from researchers, developers, and cyberinfrastructure experts as well as the Parsl team. Save ParslFest in your calendar!

If you are interested in attending or presenting at the (free) ParslFest meeting, please complete the registration form.

Parsl featured in Research Software Encyclopedia.

Braid project is funded by Department of Energy (DOE).

Globus Labs will be speaking at the 2020 DOE Data Days (D3) virtual event about Gladier: An Architecture to Enable Modular Automation of Data Capture, Storage, and Analysis at Experimental Facilities (presentation link forthcoming).

Aarthi Koripelly has two posters accepted in ACM Student Research Competition. Congratulations!

Timothy Shaffer has joined Globus Labs as a research intern! Tim will be working on managing Python software environments in Parsl to increase application portability across sites.

The funcX project website is now launched.

Ganesh Sivaraman’s paper, entitled “Electrically Sensing Hachimoji DNA nucleotides through a hybrid graphene/h-BN nanopore”, has been published to the journal Nanoscale.

The funcX project is seeking skilled Python developers to create and operate the funcX platform. funcX is designed to enable managed function execution across a distributed ecosystem of function execution endpoints deployed on heterogeneous computers: from laptops to supercomputers.

‘DLHub: Simplifying Publication, Discovery, and Use of Machine Learning Models in Science’, led by Zhuozhao Li, has been accepted for publication by the Journal of Parallel and Distributed Computing.

Kyle has received funding from the Exascale Computing Project to investigate Exascale workflow needs as part of the ExaWorks project

Ian has received funding from the Exascale Computing Project to explore Exascale data management needs

Kyle Chard presented “OAuth SSH with Globus Auth” and Rick Wagner presented “An Open System for Pervasive Use of Persistent Identifiers” at PEARC 2020.

funcX and Ryan Chard featured in Argonne’s ALCF Newsletter about using supercomputers to conduct real-time analysis of COVID-19 proteins.

Ganesh Sivaraman’s paper, entitled “Machine-Learned interatomic potentials by active learning: amorphous and liquid hafnium dioxide”, has been published to the Nature Partner Journal on Computational Materials.

“Real-time HEP analysis with funcX, a high-performance platform for function as a service”, led by Anna, has been accepted for publication to CHEP 2019 - 24rd International Conference on Computing in High Energy and Nuclear Physics proceedings.

Rohan Kumar successfully defended his Master’s thesis, entitled “Coding the Computing Continuum: Enabling Fluid Execution of Functions in Heterogeneous Computing Environments” .

Logan’s paper, titled “Quantum Chemically Informed Machine Learning: Prediction of Energies of Organic Molecules with 10 to 14 Non-Hydrogen Atoms”, is accepted to the Journal of Physical Chemistry A.

Logan’s paper, titled “A High-Throughput Structural and Electrochemical Study of Metallic Glass Formation in Ni-Ti-Al”, is accepted to ACS Combinatorial Science.

Marcus Schwarting gave a presentation to summer students at Argonne National Lab on Machine Learning for Materials Science and Chemistry.

2020 CDAC Summer Lab has started with Globus Labs hosting four students.

Maksim Levental has joined Globus Labs as a PhD student! Max earned his BS in Math from Florida State University, and his MS in Computer Science from University of Florida. He will be working on problems at the intersection of Machine Learning and Distributed HPC. With this next step taken, Max will finally no longer be a “Florida man” (not that it’s considered bad to be a “Florida man”, unless you wind up in the news). Max is still early in his career, though, so anything is possible.

Ben, Hong, Marcus, Tyler presented at CHiMaD: Center for Hierarchical Materials Design (presentation link forthcoming).

Targeting SARS-CoV-2 with AI- and HPC-enabled Lead Generation: A First Data Release is now available on arXiv.

Logan’s paper, titled “The MolSSI QCArchive Project: An Open-Source Platform to Compute, Organize, and Share Quantum Chemistry Data”, is accepted to WIREs Computational Molecular Science.

Globus Labs is contributing to the fight against COVID-19 and making our research data publicly available. Check out our data release here (DOI: 10.26311/LIT).

Ian and Kyle received an NSF Award to support our funcX project.

Ian and Kyle have 2 papers accepted to PEARC20, one titled “OAuth SSH with Globus Auth”, and one titled “An Open Ecosystem for Pervasive Use of Persistent Identifiers”.

Kyle, Ian, and Andrew Chien presented QoS-Aware Distributed Serverless Computing at the annual CERES summit.

Zhuozhao Li presented work at UC Merced.

Ian’s paper titled: A regional nuclear conflict would compromise global food security is accepted to PNAS.

Hong’s paper titled: “SciNER: Extracting Named Entities From Scientic Literature” is accepted to ICCS20.

Our paper titled: “funcX: A Federated Function Serving Fabric for Science” has been accepted to HPDC20.

Globus Labs is contributing to the fight against Covid-19. Read more here.

Ryan Chard presented 5G Enabled Energy Innovation Workshop (5GEEIW) at Westin Hotel in Chicago, IL. ``

Kyle Chard presented update of complex workflow in HEP-CCE at the Argonne National Laboratory, Chicago, IL.

High school and undergraduate students from the Greater Chicago Area are invited to apply for the CDAC Data and Computing Summer Lab. Applications due February 28th!

Ian Foster is presenting Automating the Research Data Lifecycle with Automate at eResearch 2020 in Dunedin City, NZ.

Logan Ward is part of the ExaLearn Tutorial: Co-Design Center for Exascale Machine Learning Technologies at the Exascale Computing Project’s Annual Meeting in Houston, TX.

Ben received an ARPA-E award to partner with researchers at MIT (led by Gomez-Bombarelli) and UTRC to discover new multicomponent oxide catalyst materials using machine learning approaches.

Yadu Babuji presented How funcX uses k8s at the Chicagoland K8s-HEP Meetup in Chicago, IL.

Kyle Chard Presented Parsl: Pervasive Parallel Programming in Python at the January Chicago PyData Meetup.

Ryan Chard presented on “Cyberinfrastructure for Autonomous Science” at Argonne National Lab’s Advanced Photon Source AI/ML Workshop. (https://www.anl.gov/article/advanced-photon-source-aiml-workshop).

Ian Foster penned a Globus blog post Globus Turns 10^18 about Globus history, a review of 2019, and Globus being projected to reach 1 exabyte of data transferred in 2020.

Ian Foster featured in The Big Questions: Ian Foster on High-Performance Computing by the U.S. Department of Energy’s Office of Science blog.

Ian Foster presents AI for Science at Argonne as part of Argonne National Laboratory’s Computational Science Seminar Series.

Ian Foster presented Mechanisms Empowering Discovery at the U.S. Department of Energy’s Advanced Scientific Computing Advisory Committee’s January meeting.

Takuya Kurihana presented Developing Unsupervised Learning Models for Cloud Classification and Cloud Characterization with Deep Learning II at the AGU Fall Meeting in San Francisco, CA.

Ian Foster presented ParaOpt: Automated Application Parameterization and Optimization for the Cloud on behalf of former Globus Labs Master’s students Chaofeng Wu and Ted Summer at CloudCom 2019 in Sydney, Australia.

Tyler Skluzacek presented Serverless Workflows for Indexing Large Scientific Data (WoSC ‘19) and Dredging a Data Lake: Decentralized Metadata Extraction (Doctoral Symposium) at Middleware ‘19 in Davis, CA.

Zhuozhao Li presented a talk Parsl: Pervasive Parallel Programming in Python at Python for High Performance and Scientific Computing for Birds of a Feather on SC 2019.

Logan Ward gave a presentation titled “Materials Data and AI Efforts in the US” on Materials Discovery Meets AI: Current Progress and Intensive Hands-on in Australia.

Arjun Shankar presented a Demo about using Parsl to coordinate multi-site runs in DCDE project at the DOE booth on SC 2019.

Check out Ian’s latest paper, “Scientific Image Restoration Anywhere” published on 1st Annual Workshop on Large-Scale Experiment-in-the-Loop-Computing.

Check out Ian’s latest paper, “A Codesign Framework for Online Data Analysis and Reduction” published on The 14th Workshop on Workflows in Support of Large-Scale Science.

Check out Ian’s latest paper, “MemXCT: Memory-Centric X-Ray CT Reconstruction with Massive Parallelization” published on IEEE/ACM Supercomputing (SC19) Conference on High Performance Computing Networking, Storage and Analysis.

Check out Ian’s latest paper, “Estimation of RTT and Loss Rate of Wide-Area Connections Using MPI Measurements” published on Innovating the Network for Data Intensive Science.

Check out Ian’s latest paper, “Deep Learning Accelerated Light Source Experiments” published on Deep Learning on Supercomputers.

Check out Ian’s latest paper, “Aggregating Local Storage for Scalable Deep Learning I/O” published on Deep Learning on Supercompusters.

Ian Foster co-organized The 5th International Workshop on Data Analysis and Reduction for Big Scientific Data, in conjunction with SC19.

Ian Foster co-organized the Deep Learning on Supercomputers Workshop, in conjunction with SC19.

Matthew E. Baughman’s poster titled “Enhancing Neural Architecture Search with Speciation and Inter-Epoch Crossover” was accepted to ACM Student Research Competition on SC 2019.

Kathryn Leung’s poster titled “Walking the Cost-Accuracy Tightrope: Balancing Trade-Offs in Data-Intensive Genomics” was accepted to ACM Student Research Competition on SC 2019.

Zhi Hong passed his Master defense titled “Enabling Generalizable Scientific Named Entity Recognition” on November 8th. Congratulations Hong!

Anna Woodard presented the talk Real-time HEP analysis with funcX, a high-performance platform for function as a service at CHEP2019.

Matt Baughman’s paper titled “Measuring, Quantifying, and Predicting the Cost-Accuracy Tradeoff” was accepted to The Third IEEE International Workshop on Benchmarking, Performance Tuning and Optimization for Big Data Applications (BPOD 2019).

Check out Kyle’s latest paper, “Enabling real-time multi-messenger astrophysics discoveries with deep learning” published on Nature Review Physics.

Tyler SKluzacek’s paper titled “Serverless Workflows for Indexing Large Scientific Data” was accepted to 5th Workshop on Serverless Computing (WoSC ‘19).

Tyler SKluzacek’s paper titled “Dredging a Data Lake: Decentralized Metadata Extraction” was accepted to Middleware ’19 Doctoral Symposium.

Takuya Kurihana’s abstract titled “Cloud Characterization With Deep Learning II” was accepted to American Geophysical Union (AGU) 2019 Fall Meeting.

Congratulations to Ian Foster for being honored as one of five DOE Office of Science Distinguished Scientists Fellows! The Department of Energy will provide $1 million in funding to support Ian’s pioneering research on data, computing and artificial intelligence.

Kyle Chard gave a talk titled “Globus Research Data Management Service: Capabilities, International Experiences, and Future Plans” on eResearch Australasia 2019 in Australia.

Globus labs projects are nominated in several categories in this year’s HPCwire Readers’ Choice Awards. Vote for us in categories: “Best Use of HPC in Physical Sciences”, “Best HPC Programming Tool or Technology”, “Best HPC Storage Product or Technology”

ParslFest, the first Parsl Community Meeting will be held October 3-4, 2019 at the University of Chicago. The two day meeting included presentations and discussions on the first day and tutorials, advanced topics, and helpdesk sessions on the second day.

Ian served as a panelist on the topic of “Engaging Male Allies” at Women in HPC: Diversifying the HPC Community and Engaging Male Allies.

Kyle presented “Serverless Science for Simple, Scalable, and Shareable Scholarship” at eScience 2019.

Kyle served as a panelist on the topic of “Effective Checklists for Developers and Researchers” at Gateways 2019.

Kyle’s paper “Application of BagIt-Serialized Research Object Bundles for Packaging and Re-execution of Computational Analyses” was presented at the Research Objects workshop at eScience 2019.

Kyle presented “Big Data Bags: A Scalable Packaging Format for Science” at the Research Objects workshop at eScience 2019.

Kyle presented “Active Learning Yields Better Training Data for Scientific Named Entity Recognition” at eScience 2019.

Ben received an NSF award, jointly with the University of Wisconsin-Madison, to enable continued research towards building a materials innovation infrastructure compatible with modern machine learning techniques.

Check out Ben and Ian’s new paper, “Virtual Excited State Reference for the Discovery of Electronic Materials Database (VERDE Materials DB): An open-access resource for ground and excited state properties of organic molecules”.

Matt (Enhancing Neural Architecture Search with Speciation and Inter-Epoch Crossover) and Kathryn’s (Walking the Cost-Accuracy Tightrope: Balancing Trade-offs in Data Intensive Genomics) posters were accepted to SC2019.

Kyle Chard’s paper: Tales at Scale: Toward Enabling Reproducibility for Data-Intensive Research using the Whole Tale Platform presented at ParCO2019.

Ian gives a keynote at ParCO2019 on Coding the Continuum.

Ben Blaiszik, Marcus Schwarting, and Ian Foster receive Argonne LDRD award to prototype the development of automated laboratories coupled with AI.

Dr. Roselyne Tchoua, 2019 Globus Labs Alumna, starts as Assistant Professor at the School of Computing at DePaul University (Chicago, IL). Congratulations, Roselyne!

The CDAC summer internship officially concludes this week! Special thanks to Aarthi, Avery, Katy, Nikki, Ruby, Ryan, Sydney, Teresa, and Tobias for their contributions to Globus Labs research projects!

The inaugural ParslFest Community meeting will take place Oct. 3rd-4th at the University of Chicago. See the Parsl website for more information.

Check out Logan’s, Ben’s, and Ian’s new paper, Machine Learning Prediction of Accurate Atomization Energies of Organic Molecules from Low-Fidelity Quantum Chemical Calculations, which was accepted for publication in MRS Communications Special Issue on AI!

Takuya Kurihana’s paper “Cloud Characterization with Deep Learning” accepted to Climate Informatics 2019.

Ryan Chard previews Globus Automate at the September Campus Research Computing Consortium meeting.

Kyle Chard joined Argonne Physicist Katrin Heitmann to discuss “Simulating and Sharing the Universe”. You can see the full interview on Facebook.

Globus Labs project funcX featured in HPC-Wire. Check it out here!

Logan Ward taught a short coarse on using convolutional neural networks for scientific research at the ALCF AI for Science Tutorial.

Kyle Chard presents “Serverless Supercomputing: High Performance Function as a Service” at Aalto University in Otaniemi, Finland.

Serverless Supercomputing: High Performance Function as a Service for Science paper is now available on arXiv.

Ravi Madduri, Kyle Chard, and Ian Foster publish “Reproducible big data science: A case study in continuous FAIRness” in PLOS ONE. Check it out here!

Zhuozhao Li presented Publishing and Serving Machine Learning Models with DLHub at PEARC 2019.

Zhuozhao Li presented Publishing and Serving Machine Learning Models with DLHub at PEARC 2019.

Rick Wagner presented Petrel: A Programmatically Accessible Research Data Service at PEARC 2019.

Kyle Chard presented Scalable Parallel Programming in Python with Parsl at PEARC 2019.

Anna Woodard will present a talk titled “Real-time HEP analysis with funcX, a high-performance platform for function as a service” at CHEP 2019.

Check out our new paper, A Data Ecosystem to Support Machine Learning in Materials Science which was accepted for publication in MRS Communications Special Issue on AI!

Roselyne Tchoua’s paper “Active Learning Yields Better Training Data for Scientific Named Entity Recognition” is accepted at the IEEE 15th International Conference on eScience, 2019.

Ian Foster presented on AI at Scale for Materials and Chemistry at the Department of Energy AI for Science Town Hall hosted at Argonne National Laboratory.

Logan Ward presented a tutorial on Supervised Learning (regression, classification, CNN) at the AI for Science Student workshop at Argonne National Laboratory.

Logan Ward presented a talk on “Materials Design” at the AI for Science Student workshop hosted at Argonne National Laboratory.

Ben Blaiszik presented a talk on A FAIR Approach to Publishing and Sharing Machine Learning Models at the AI for Science Student workshop at Argonne National Laboratory.

Roselyne Tchoua will present her PhD dissertation on July 2nd from 11AM at John Crerar Library (JCL) 298. Her dissertation title is HYBRID HUMAN-MACHINE SCIENTIFIC INFORMATION EXTRACTION FRAMEWORKS. Go Roselyne !

Yadu Babuji presented Parsl talk at the HPDC 2019 on June 26th. His presentation title is Parsl: Pervasive Parallel Programming in Python.

Matt Baughman presented his AWS talk at the ScienceCloud Workshop on June 25th. He presented Deconstructing the 2017 Changes to AWS Spot Market Pricing.

Ian Foster gave a keynote at the ScienceCloud Workshop on June 25th. His present title is Coding the Continuum.

Matt Baughman After working with Globus Labs part-time for nearly two years, Matt Baughman will be joining GLabs full time as a PhD student. Matt earned his Bachelor in Computational Sciences from the Minerva Schools at KGI where he studied computational analysis and theory. He will continue working with Kyle Chard on the development of automated optimization approaches for scientific cloud computing.

Roselyne Tchoua presented “Creating Training Data for Scientific Named Entity Recognition with Minimal Human Effort” at ICCS 2019 in Faro, Algarve, Portugal.

Takuya Kurihana presented his cloud model at the CS-1 Workshop on June 8th. He proposed a potential application on the new ALCF deep learning machine CS-1 .

Marcus Schwarting gave a general overview of current techniques in machine learning and how they can be applied to problems in materials science. He also highlighted the use of DLHub and MDF for facilitating the machine learning model building process.

Roselyne Tchoua to present “Enabling Generalizable Machine Learned Scientific Named Entity Recognition” at SIGSIM PADS 2019 on Jun. 4 in Chicago, IL.

Dr. Ian Foster, recipient of the 2019 IEEE Computer Society Charles Babbage Award, presents keynote entitled “Coding in the Continuum” at IPDPS 2019 in Rio de Janeiro, Brazil.

Globus Labs has three papers on DLHub, Parsl, and Petrel accepted to PEARC19.

Dr. Kyle Chard to present “FuncX: High Performance Function-as-a-Service for Science” at the 11th International Workshop on Science Gateways in Ljubljana, Slovenia (Jun. 12-14).

Kyle Chard chaired and Zhuozhao Li vice-chaired the Workshops Proceedings Committee at IPDPS 2019.

Logan Ward co-organized a symposium at the MRS Spring Meeting integrating AI with materials engineering.

Globus Labs presents posters on DLHub, funcX, and Parsl at GCASR 2019 at DePaul University.

The annual GlobusWorld Conference is being held in Chicago, IL on May 1-2, 2019!

Kyle presented Parsl at the Enabling Multi-Messenger Astrophysics in the Big Data Era Workshop.

Our paper Reproducible big data science: A case study in continuous FAIRness was published in PLOS One.

Roselyne B. Tchoua was featured in MediaPlanet’s Women in Computing Campaign. You can find the article online here: Advice From Women in Computer Sciences”.

Roselyne B. Tchoua, Zhi Hong, Logan T. Ward, Kyle Chard, and Ian T. Foster publish paper, “Creating Training Data for Scientific Named Entity Recognition with Minimal Human Effort”, at ICCS 2019.

Sandro Fiore (Ph.D) has joined the U. Chicago and Globus Labs as a Visiting Scientist. He is Senior Scientist and Head of the Data Science and Learning Research Team at the Euro-Mediterranean Center on Climate Change (CMCC). His research focuses on scientific data management, in particular high performance data analytics workflows as well as provenance and computational reproducibility aspects.

Globus Labs will be giving a talk on Parsl at the 28th International ACM Symposium on High-Performance Parallel and Distributed Computing (HPDC’19)

Ryan Chard presented FuncX: A Function Serving Platform for HPC at the Singularity User Group Meeting.

Logan Ward is invited to give a talk on Machine-learning-aided Design of Metallic Glasses at TMS 2019.

The NIST Center for Hierarchical Materials (CHiMaD) has been extended for an additional 5 years. We are excited to continue expanding on the solid foundation that CHiMaD hasbuilt. Globus Labs projects including the Materials Data Facility and the Polymer Property Predictor and Database (PPPDB) are funded through CHiMaD. [Full Report]

Ian Foster named recipient of the 2019 IEEE Computer Society Charles Babbage Award! Congratulations!

Join our UChicago Data and Computing Summer Lab. Applications due by March 10th, 2019.

Ryan Chard presents DLHub at CANDLE Workshop.

Ian Foster presents Infrastructure for data-driven discovery: Materials Data Facility and DLHub at nano manufacturing (nanoMFG).

Greg Pauloski’s paper “Deep Neural Network Training with Distributed K-FAC” accepted for publication to TPDS.

Marcus Schwarting presents Machine Learning Approaches for XRD Analysis of the HTEM Database at the APS User Science Seminar.

Logan Ward presents Combining Machine Learning and High-Throughput Science at International Symposium on Materials Informatics.

Ricardo B. Lourenço will present Cloud Characterization with Deep Learning at the University of São Paulo on March 28th and the University of Campinas on March 29th.

Kyle Chard’s paper on Cloud Futurology accepted for publication to IEEE Computer 2019.

Ian Foster presents Globus Automate and Globus BoF at eResearch 2019 in Auckland, NZ.

Ian Foster gives a keynote address at Multicore World 2019 entitled “Learning Systems for Science”.

Ricardo Lourenco received an Outstanding Student Paper Award from the American Geophysical Union for his work titled Cloud Characterization with Deep Learning.

Globus Labs will be giving a talk on DLHub at the 2019 IEEE International Parallel & Distributed Processing Symposium (IPDPS)

Matt Baughman, Ryan Chard, Logan Ward, Jason Pitt, Kyle Chard and Ian Foster publish paper, “Profiling and Predicting Application Performance on the Cloud”, at UCC 2018.

Ricardo B. Lourenço and Takuya Kurihana presented a poster at the AGU fall meeting 2018 on understanding Cloud characterization with deep learning

Swift/T, a dataflow programming framework for scientific supercomputing workflows developed in part by Ian Foster, wins a R&D 100 Award.

Tyler Skluzacek presented his Master’s Thesis, “Automated Workflows for Extracting and Deriving Metadata from Disorganized Data Swamps”.

Globus has a great turn out at their booth at the Supercomputing Conference (SC18). You can learn more about it here.

Globus Lab’s summer 2018 REU students Luann Jung and Brendan Whitaker win Undergraduate Award at Supercomputing 2018 for their work on “data swamps”.

Zhi Hong received the “Volunteer of the day” award while volunteering at the SC18 conference.

Kyle Chard presented Parsl’s use in the LSST DESC project to simulate LSST images at the PyHPC workshop.

Kyle Chard presented “Reproducible big data science: A case study in continuous FAIRness” at Research Object Workshop 2018.

Marcus Schwarting joined the Globus Labs Team. Marcus is working on machine learning methods for materials discovery and testing. He will also be working on a project 3M on leveraging machine learning for CFD calculations. He is working on a master’s degree in computer science with an emphasis on machine learning from Georgia Tech.

Logan Ward gave a talk on Materials Database at the Workshop on “Gap Analysis: Materials Discovery through Data Science at Advanced User Light Sources” in Santa Fe, New Mexico.

Yuliana Zamora presented a poster at CERES titled “Data Driven Analysis of GPU Benchmarking”. The poster describes a data-driven framework that makes use of data-analysis tools, such as machine learning and transfer learning, to reduce the time needed for engineers and researchers to predict the performance of a given software kernel on an arbitrary GPU architecture.

Roselyne Tchoua presented a poster at CERES titled “Towards Hybrid Human-Machine Scientific Information Extraction”. The poster describes different combinations of human-machine approaches to extracting scientific facts from the literature, specifically polymer names and properties from a relevant scientific journal in polymer sciences.

Ian attended the 2018 Workshop on Clusters, Clouds, and Data for Scientific Computing (CCDSC) in Lyon, France. He presented on the computer systems challenges raised by the rise of deep learning in science, and technologies that Globus Labs is developing to address those challenges.

The Whole Tale team is holding a Workshop on Tools and Approaches for Publishing Reproducible Research on September 13-14, 2018 at the Big Ten Center in Chicago.

Ian Foster attended JupyterCon, The Official Jupyter Conference. He presented a talk about Scaling collaborative data science with Globus and Jupyter.

Tyler Skluzacek will be giving a talk at eScience 2018, this October in Amsterdam for his accepted paper: Skluma: An extensible metadata extraction pipeline for disorganized data.

Takuya Kurihana has joined Globus Labs as a PhD student! Takuya earned his BS in Meteorology from University of Tsukuba, where he studied data assimilation for weather forecasting application. He will be working with Liz Moyer in Geophysical science department to apply Deep Learning for the development of climate model.

Logan Ward presented at the NIST Workshop on Artificial Intelligence for Materials Science (AIMS) on how machine learning can be combined with atomic scale calculations.

Roselyne Tchoua gave an invited talk for a poster submitted to the 2018 New York Scientific Data Summit (NYSDS) about how to use word embedding to generate training data for scientific named entity recognition.

Anna Woodard attended Computing in High Energy and Nuclear Physics (CHEP) 2018. She presented a talk on Interactive, scalable, reproducible data analysis with containers, Jupyter, and Parsl.

Kyle Chard presented Parsl at the Data Visualization and Exploration in the LSST Era workshop held at NCSA, the University of Illinois.

Ricardo Barros Lourenco presented “Atmospheric Cloud Characterization with Deep Learning on Google Earth Engine” at Google Earth Engine User Summit 2018.

High-Throughput Neuroanatomy and Trigger-Action Programming: A Case Study in Research Automation. 1st International Workshop on Autonomous Infrastructure for Science.

Logan Ward had his paper “Matminer: An open source toolkit for materials data mining” published in Computational Materials Science Journal.

Logan Ward gives talk on “machine learning and materials modeling” at CECAM workshop

Ian Foster and Kyle Chard had their paper “Predicting Amazon Spot Prices with LSTM Networks” accepted at ScienceCloud ‘18

Kyle Chard gave a presentation entitled “Parsl: A Parallel Scripting Library for Python” at Notre Dame University

Globus Labs joined past and present researchers of the Computation Institute to discuss how the accomplishments of the institute since 1999 teach us about the next 20 years of computational research at UChicago, Argonne, and beyond.

A full webcast of Logan Ward’s talk at the International Workshop on Machine Learning for Materials Science is available online

Globus Labs slides from Ian Foster’s Globus World 2018 keynote are available online

Kyle Chard will be giving a presentation about Parsl - a parallel scripting library for Python - at Scipy 2018 this July in Austin. We hope to see you there!

Ian Foster gave a keynote presentation about “Going Smart and Deep on Materials” at the Annual Meeting of The Minerals, Metals, and Materials Society (TMS) society. There, he talked about our work towards making data science more prevalent in materials engineering.

Anna Woodard has joined Globus Labs as a postdoc! Anna earned her PhD in Physics from Notre Dame, where she studied top quarks as part of the CMS collaboration. She will be working with Kyle Chard to simplify parallel computing in Python as part of the Parsl Project.

BigDataX Research Experiences for Undergraduates funded by NSF. More info here.

Kyle Chard gave a presentation at UIUC titled “Parsl: A Parallel Scripting Library for Python”. Learn more about Parsl here.

Check out Ian Foster’s post on his New Year’s Relections on Globus.

SRC: Assessing Verticalization Effects on Urban Safety Perception. ACM SIGSPATIAL Special, 2018.

Ian Foster and Logan Ward attended a Deep Learning Workshop at Argonne National Laboratory on Jan. 16, 2018. Ian presented about deep learning at ANL as the new lead of the Data Science and Learning Division; Logan presented about deep learning and materials design.

Ian Foster attended the 24th IEEE International Conference on High Performance Computing, Data and Analytics (HiPC) in Jaipur, India. He presented a keynote talk entitled “Computing Just What You Need: Online Data Analysis and Reduction at Extreme Scales” on Dec. 21, 2017.

Ian Foster has been appointed to lead Argonne National Laboratory’s new Data Science and Learning Division on Nov. 08, 2017.

Ian Foster has presented a demo on a service that integrates data from the Materials Data Facility to Argonne Leadership Computing Facility, using Globus, at the SC’17 Supercomputing Conference in Denver, CO on Nov. 14, 2017.

Tyler Skluzacek and Yuliana Zamora received honors at the 2017 annual CERES summit at the University of Chicago on Sept. 27, 2017.

RIPPLE: Home Automation for Research Data Management. First International Workshop on Serverless Computing (WoSC), 2017.

Skluma: A statistical learning pipeline for taming unkempt data repositories. 29th International Conference on Scientific and Statistical Database Management (SSDBM), 2017.

Ben Blaiszik, Kyle Chard, Ryan Chard, and Tyler Skluzacek present their research. (GlobusWorld), 2017

Explaining Wide Area Data Transfer Performance. ACM International Symposium on High-Performance Parallel and Distributed Computing (HPDC), 2017.

ATLAS: A visualization and analysis framework for geospatial datasets. XVIII Brazilian Symposium on Remote Sensing. May 28-31, 2017.

Our paper Experimental Analysis of File Transfer Rates Over Wide-Area Dedicated Connections received the best paper award at the 18th IEEE International Conference on High Performance Computing and Communications, 2016.

Our paper Globus Auth: A Research Identity and Access Management Platform received a best of conference award at the IEEE 12th International Conference on eScience, 2016.

William Agnew received the first place award in the ACM Student Research Competition for his paper titled Touring Dataland? Automated Recommendations for the Big Data Traveler.

Klimatic: A Virtual Data Lake for Harvesting and Distribution of Geospatial Data. 1st Joint International Workshop on Parallel Data Storage & Data Intensive Scalable Computing Systems (PDSW-DISCS), 2016.

Kyle Chard received the IEEE Computer Science Technical Committee on High Performance Computing (HPC) Award for Excellence for Early Career Researchers in HPC.

We have several Postdoc and Ph.D. student positions available. Join Us.

A Hybrid Human-computer Approach to the Extraction of Scientific Facts from the Literature. Procedia Computer Science 80, 386-397.

Globus Auth: A Research Identity and Access Management Platform. Ian Foster. eScience 2016, Baltimore MD, Oct 25.