September 20, 2019

Systems for Machine Learning request for proposals

This Research Award is now closed

About

Machine learning (ML) methods have been widely applied to a large variety of day-to-day services hosted by Facebook. These services include personal recommendation, content understanding (language translation, vision, and speech), integrity, and more. As Facebook increasingly relies on ML systems to better address the individual needs of users, we also invest in full-stack and end-to-end optimization in our infrastructure to guarantee the quality of services in performance, availability, accuracy, and reliability. Such high QoS is achieved holistically across the entire computing stack from robust machine learning development frameworks down to maximizing compute/power efficiency in data centers. The breadth of this endeavor is often beyond internal research and development, and includes outreach to and collaboration with research communities in academia.

To sustain traditional computing at scale for ever-growing machine learning workloads in our data centers, we continue to call for impactful solutions in the areas of developer toolkits, compilers/code generation, system architecture, memory technologies, and ML accelerator support. In addition, emerging issues around compute trustworthiness and data privacy have become forthcoming challenges and require interdisciplinary innovation between system engineers and privacy/security scientists. We are looking forward to novel system design approaches to enable privacy-preserving machine learning, differential privacy, or multi-party compute. Finally, we are interested in exploring any disruptive, game-changing technology for the landscape of training and inference for machine learning.

To foster further innovation and to deepen our collaboration with academia, Facebook is pleased to invite faculty to respond to this call for research proposals pertaining to the aforementioned topics. Applicants should submit a maximum two-page proposal detailing what contribution their research is expected to make, how the research domain will benefit from the work, and a budget overview of how the proposed funding will be used. Proposals are highly encouraged to focus funding on project personnel, especially PhD students. Proposals from small interdisciplinary teams with complementary expertise are also encouraged. We anticipate awarding six to nine proposals, in the range of $50,000 to $100,000 with larger awards for interdisciplinary collaborations between multiple PIs. Payment will be made to the proposer’s host university as an unrestricted gift.

Applications are now closed

Application Dates

Notifications will be sent by email to selected applicants by December 2019.

  • Launch Date September 20, 2019
  • Deadline
  • Winners Announced December 2019

Areas of Interest

This year, we are particularly interested in proposals that fall into these categories:

  • Scalable, elastic, and reliable distributed machine learning and inference
  • System and architecture support for personalized recommendation systems
  • Programming language and compilers for platform-agnostic machine learning
  • Resource provisioning for efficient inference/training in heterogeneous data centers
  • On-device training and inference
  • System and architecture support for privacy-preserving machine learning
  • System support for multi-party compute and private/secure inference
  • Emerging technologies, such as near-memory processing and in-memory computing systems applied to machine learning
  • Novel machine learning systems beyond neural networks
  • Emerging technologies for efficient machine learning

Requirements

Proposals should include

  • A summary of the project (1-2 pages) explaining the area of focus, a description of techniques, any relevant prior work, and a timeline with milestones and expected outcomes
  • A draft budget description (1 page) including an approximate cost of the award and explanation of how funds would be spent
  • Curriculum Vitae for all project participants
  • Organization details. This will include tax information and administrative contact details

Eligibility

  • Awards must comply with applicable U.S. and international laws, regulations, and policies.
  • Applicants must be current full-time faculty at an accredited academic institution that awards research degrees to PhD students.
  • Applicants must be the Principal Investigator on any resulting award.

Timing and dates

  • Applications are now closed.
  • Notifications will be sent by email to selected applicants by December 2019.

Frequently Asked Questions

Most of the RFP awards are an unrestricted gift. Because of its nature, salary/headcount could be included as part of the budget presented for the RFP. Since the award/gift is paid to the university, they will be able to allocate the funds to that winning project and have the freedom to use as they need. All Facebook teams are different and have different expectations concerning deliverables, timing, etc. Long story short – yes, money for salary/headcount can be included. It’s up to the reviewing team to determine if the percentage spend is reasonable and how that relates to the decision if the project is a winner or not.

We are flexible, but ideally proposals submitted are single-spaced, Times New Roman, 12 pt font.

Research awards are given year-round and funding years/duration can vary by proposal.

Yes, award funds can be used to cover a researcher’s salary.

Budgets can vary by institution and geography, but overall research funds ideally cover the following: graduate or post-graduate students’ employment/tuition; other research costs (e.g., equipment, laptops, incidental costs); travel associated with the research (conferences, workshops, summits, etc.); overhead for research gifts is limited to 5%.

One person will need to be the primary PI (i.e., the submitter that will receive all email notifications); however, you’ll be given the opportunity to list collaborators/co-PIs in the submission form. Please note in your budget breakdown how the funds should be dispersed amongst PIs.