BY GINA MANTICA
The Boston University (BU) Rafik B. Hariri Institute for Computing and Computational Science & Engineering: Red Hat Collaboratory is excited to announce the first recipients of the Red Hat Collaboratory Research Incubation Award! Through BU and Red Hat’s expanded partnership, the Red Hat Collaboratory seeks to create more trustworthy, reliable, scalable, self-operating, distributed, heterogeneous compute platforms that stretch from edge devices to cloud datacenters. “This collaborative model gives us the opportunity to increase the diversity and richness of open engineering and operations projects we undertake together, and also allows us to pursue fundamental research under one umbrella,” said Heidi Picher Dempsey, Research Director of the Northeast U.S. for Red Hat.
The Red Hat Collaboratory also enables innovative partnerships between the academic research community and open source community. Collaboration in systems research at this scale can have a profound impact on society. “Collaboration not only enables research projects that have an impact, but also high enough quality research artifacts upon which other research projects can build,” said Orran Krieger, Co-Director of the Red Hat Collaboratory.
Projects funded through the Red Hat Collaboratory Research Incubation Award are open source and focus on problems of distributed, operating, security, or network systems whose solutions show promise for advancing their fields and impacting the tech industry. The Red Hat Collaboratory has awarded 16 projects from BU faculty members and industry collaborators, totaling more than $2.3 million in funding. “The diverse collection of proposals we received promised to carve out a wide impact in the many research and societal efforts that rely upon systems research,” said Ari Trachtenberg, a Professor of Electrical & Computer Engineering at BU who was instrumental in developing the funding program.
Learn more about the funded projects below.
AI for Cloud Ops
BU faculty members Ayse Coskun, Alan Liu, and Gianluca Stringhini; Red Hat researchers Steven Huels, Marcel Hild, and Daniel Riek; and IBM researcher Fabio Oliviera will pursue work at the intersection of artificial intelligence (AI) and cloud systems through a two-year funded project. The researchers intend to create new methods for fusing and representing systems data to enable AI-based analytics and will build, apply, and scale AI frameworks to improve performance, management, security, compliance, and resilience problems in the cloud. “We aim to help deliver easily-accessible and open source AI technologies that will cater to both developers and administrators solving real-world performance, resilience, and security challenges,” the team wrote.
Creating a global open research platform to better understand social sustainability using data from a real-life smart village
BU faculty members Christos Cassandras, Vasiliki Kalavri, John Liagouris, and Mayank Varia; and Red Hat researchers Alexandra Machado, Jim Craig, and Christopher Tate, will team up with SmartaByar, an organization specializing in Smart Villages, to create a global, open research platform where researchers can collaborate to define a link between well-being and eco-smart cities. To study social sustainability, the researchers’ platform will leverage real-life data from Veberöd, Sweden in a virtual “village.” This enables the researchers to test how different technologies might impact the real village. “The unique opportunity of this collaboration is to leverage the strengths of all the cross-sector stakeholders in order to have a global open research platform where all data, findings and studies can be correlated and shared for a better understanding of what smart social sustainability means,” the team wrote.
DISL: A Dynamic Infrastructure Services Layer for Reconfigurable Hardware
BU faculty member Martin Herbordt will work with Red Hat researchers Uli Drepper and Ahmed Sanaullah to create a generic operating system and firmware for reconfigurable hardware called Dynamic Infrastructure Services Layer, or DISL. DISL decouples the device specific hardware, operating systems, and application spaces to enable substantially higher productivity through reduced developer effort, well-defined developer roles, standard software interfaces, and greater code compatibility across chips and vendors. “DISL will make it easier for researchers to focus on specific, well-defined areas of innovation instead of repeatedly reinventing the hardware stack,” the team wrote.
Towards high performance and energy efficiency in open-source stream processing
BU faculty members Vasiliki Kalavri and Jonathan Appavoo will work with Red Hat researcher Sanjay Arora to investigate the energy and performance behavior of open-source stream processing platforms. The researchers will develop an open-source Mass Open Cloud (MOC)-hosted experimental artifact based on Apache Flink and a visualization tool to collect and analyze energy-performance measurements for various stream processing workloads. “This project aims to demonstrate that energy efficiency and the myriad layers of software that go into an open-source streaming platform need not be incompatible,” the team wrote.
Near-Data Data Transformation
BU faculty members Manos Athanassoulis and Renato Mancuso will work with Red Hat researchers Uli Drepper and Ahmed Sanaullah to create a hardware-software co-design paradigm for data systems that implements near-memory processing. The approach has the potential to revolutionize data management by bridging the gap of analytical and transactional processing. This paradigm addresses the performance bottleneck caused by memory bandwidth and will allow both cloud and edge systems to efficiently handle mixed transactional and analytics data-intensive workloads with a better trade-off between bandwidth and latency. “The proposed software-hardware co-design methodology will also improve the collective understanding of new design models and resource management strategies that are possible in systems with programmable memory hierarchies,” the team wrote.
Practical Programming of FPGAs with Open Source Tools
BU faculty member Martin Herbordt will work with Red Hat researchers Uli Drepper and Ahmed Sanaullah to create an open source tool for reducing application development effort and turnaround time for Field Programmable Gate Arrays (FPGAs). This will enable both software and hardware developers to easily generate high quality custom hardware on FPGAs using high level languages instead of complex hardware description languages. “The overall goals of this project are to i) realize abstractions that facilitate development of high quality configurations for spatial computing, ii) leverage state-of-the-art techniques for code optimizations in performance oriented compilers, and iii) to do so in a manner that allows concurrent academic and industry efforts, with a well defined path for incorporating innovation from the former into the latter,” the team wrote.
Other funded projects include:
- Symbiotes: A New step in Linux’s Evolution, Jonathan Appavoo
- Foundations in Open Source Education, Jonathan Appavoo
- Linux Computational Caching, Jonathan Appavoo
- Robust Data Systems Tuning, Manos Athanassoulis and Evimaria Terzi
- Privacy-Preserving Cloud Computing using Homomorphic Encryption, Ajay Joshi
- Serverless Streaming Graph Analytics, Vasiliki Kalavri
- Secure cross-site analytics on OpenShift logs, John Liagouris
- Enabling Intelligent In-Network Computing for Cloud Systems, Alan Liu
- OSMOSIS: Open-Source Multi-Organizational Collaborative Training for Societal-Scale AI Systems, Eshed Ohn-Bar
- Intelligent Data Synchronization for Hybrid Clouds, David Starobinski
The scale of these projects is larger than ever before in the Red Hat Collaboratory’s history. “For the first time, we are able to support a large project with multiple students and multiple engineers working side by side on a big goal,” said Hugh Brock, Co-Director of the Red Hat Collaboratory, “If we’re successful, this model will not only make an impact in new learning and new technology, but also in the very fundamentals of the way research is done.”
To learn more about the Hariri Institute’s transformational research and stay up-to-date on these projects, click here to sign up for our newsletter.