Home / Projects / Call For Proposals

Call for Proposals - Spring 2022

Spring 2022 Call for Proposals

CROSS is seeking proposals for new research and incubator projects to be considered at our next Sponsors Meeting May 19, 2022.

CROSS values diversity and inclusion in our research and incubator projects. We invite students and researchers from traditionally excluded groups in tech and open source communities to apply to this Call.

Research Projects:

Faculty as well as Ph.D. students from UC Santa Cruz who will have completed their coursework by the end of Spring 2022 are encouraged to propose research projects that fit within the suggested topics listed below. Please note that all research projects must include a UCSC faculty member as an advisor. Selected projects receive funding for one graduate student researcher position (GSR) (students must have completed coursework to be funded by CROSS). Research projects should raise a fundamental research question and open a plausible path to open source software that would be widely adopted. Funding of a research project is intended to foster excellent research, to enable a student to become an open source software leader, and to support her work until graduation.

Incubator Projects:

Recent PhD graduates or students with expected graduation of Spring 2022 from any university are welcome to propose incubator projects in order to be accepted into the Open Source Incubator Fellowship. Incubator fellowship projects turn innovative research prototypes into successful open source software projects. This Fellowship is intended for postdoctoral scholars who want to grow contributor communities around their open source research prototypes and who can play the role of "open source ambassadors" at UC Santa Cruz, including through exemplifying open source project and community management practices, sharing their experiences as a mentor to students, and as a classroom instructors for courses that teach students how to productively engage in open source projects. New incubator projects are expected to start out with significant infrastructure created as part of a PhD project. The proposal should also show strong evidence that the fellow's project would meet with great interest by at least one well-established open source community.

 

Project Review:

Projects funded by CROSS are reviewed by the IAB and the Advisory Committee every six months, and continued funding is based on IAB and the Advisory Committee’s assessment that sufficient progress is being made. The project teams are also required to make all source code available under an open source software license and list their code repository in the public CROSS software catalog.

Important Dates

  • Proposals due: April 29, 2022 EXTENDED
  • Notification of proposal shortlist: May 5, 2022
  • IAB meeting and proposal presentations: May 19, 2022
  • Funding notifications: no later than June 2, 2022

Proposal Requirements

All project proposals should include:

  • Project Title

  • Background, overview & objectives

  • Expected time needed to meet objectives

  • Expected staffing needs

  • Explanation how project aligns with the strategic interest and overall work of CROSS

  • Brief explanation of the relevant experience / previous work of project lead (and other team members if applicable)

Additional requirements for incubator proposals can be found here

Application and Selection Process

To propose a new research or incubator project, send your proposal to CROSS Assistant Director Stephanie Lieggi by April 21, 2022. All submitted proposals will undergo an initial review by the CROSS UCSC Advisory Committee chaired by the CROSS Director. Based on this initial review, the most relevant projects will be shortlisted and applicants will be notified by April 30.

If a proposal is shortlisted by CROSS for further consideration, the project lead (or designate) will be required to present their proposal at the CROSS IAB meeting on May 19. After project presentations, members of the IAB provide recommendations to the CROSS Director, who then makes the final funding decisions in consultation with the UCSC Advisory Committee. Notifications of funding decisions should be provided by June 2.

All proposals should highlight how the proposal meets CROSS project criteria (see below). The funding from CROSS typically covers the cost of one graduate researcher (for research projects) or post-doc (for incubator projects) for the duration of the project.

Project Evaluation Criteria   

The CROSS IAB, the CROSS Director, and the UCSC Advisory Committee consider the following criteria when evaluating research and incubator projects:

For research projects

  • The project raises a fundamental research question;

  • The project opens a plausible path to open source software that would be widely adopted;

  • The project’s topic coincides with the suggested topics and meets the strategic objectives of CROSS.

For incubator projects

  • The project starts out with significant software infrastructure that a recent graduate created as part of their Ph.D. project;

  • There is strong evidence of community interest in the effort and that the resulting open source software prototype would be widely adopted; and

  • The project’s topic meets the strategic objectives of CROSS.

 

Suggested Topics for Research Projects(updated August 4, 2021)

Storage Systems

  • Computational storage devices, collective architectures
    • Extension of serverless functions infrastructure
      • Admission control based on what criteria
      • Requires general task model
    • Pipeline management, in-band, out-of-band, benefits of pipelining, explicit vs automatic fusing
      • Physical design management
      • Encryption and compression can/should be user-provided functions
    • Cooperative semantic-based sharding algorithms/strategies
      • Blending the need for semantically complete shards while supporting inexpensive data availability strategies such as eraser coding
    • Management of out-of-band jobs
      • Resource management, view management, continual optimizations
  • Management of tabular data
  • Resource Management
    • Backpressure within storage systems (or provide all details to the host and let it manage it)
  • Use Cases
    • Cloud VR/AR

 

Data Management

In-store Computing

  • Device interface semantics that allow for dynamic host/device associations
  • Device interface semantics that enable data consistency across devices
  • Domain-specific in-store data management and analytics such as genomics
  • Data privacy and access security
  • Data cleaning
  • Searching, data reduction, statistical modeling, transcoding/format conversion
  • In-memory processing models, including a secure method for offloading processing to smart SSDs
  • Runtime environments: open vs closed, standard stack vs special stack, security vs performance vs richness
    • Examples: WebAssembly, eBPF, other runtime environments that can leverage existing open source libraries
  • IO Models
    • Host managed vs device managed
    • Interactions with host-based IO models

 

Programmable Storage Systems (Programmability and Architecture of Storage Systems)

  • Open source API for in-store compute (programmable storage)
  • Open source tools for in-store compute
  • Compilers, libraries, debuggers
  • Tools to decompose applications for parallel processing of dispersed data
  • Leveraging communication between in-store compute devices o Combining in-store computing and external computing
  • Parallelism for machine learning
  • Barriers to adoption of in-store compute

 

Predictable Performance

  • Adaptable load balancing (especially using machine learning)
  • Networking with predictable performance and performance management
  • Storage maintenance
  • Coordinating system and device low-priority tasks
  • Open source frameworks to take advantage of Storage on disk- and Flash-based Zoned Block Devices, and “hybrid” CMR/SMR

Non-volatile Memory and Hybrid Memory Hierarchy Architectures (longer research horizon)

  • Hypervisor / OS architecture for heterogeneous memory
  • Memory disaggregation in the data center for data intensive applications. This will include memory object definition, representation, optimization, compiler support, and application acceleration
  • Self-aware data orchestration – predictive data positioning through intelligence and learning
  • Converging memory and IO using persistent memory – representative applications, key benefits, theory of operation
  • Operating Systems mechanisms to provide efficient/performant and secure access to non-volatile memory
  • Modeling and simulation of modifications to big data applications/SSD design to optimize performance/wear leveling
  • New data management models (e.g. location awareness) for managing massive amounts of data on storage media that includes non-volatile memory