To guide the discussion, we have some potential issues/questions. Please consider the most important ones to you (you do not need to answer them all!). Please consider diversity and inclusion, existing resources, and areas of flexibility at the end as well.

For Industry/Government:

  • Where are VLSI jobs within/outside the US, and what trends do we see at those companies?
  • What is needed to bring more of these jobs to the US?
  • Where and how do you recruit and why?
  • How do you utilize internships, academic research funding, and on-the-job training?
  • How important is it for students to have specific components of VLSI education?
    • Intro and advanced courses
    • Device manufacturing/modeling
    • Experience with contemporary processes
    • Commercial CAD tool experience
    • Team projects/large scale projects
    • Fabrication and/or testing experience
    • Scripting/Unix/Linux/System Administration
  • What is missing from VLSI education?
  • How can open-source hardware and EDA tools help?
  • What are the problems with open-source hardware and EDA?
  • How important is academic research to you?
  • How can we provide support for start-ups? What incubators are available?

For Academia:

  • How can EDA infrastructure be low cost and low maintenance?
    • Can you use Virtual Machine or Docker images?
    • Is licensing a problem? Floating licenses vs. fixed-node vs. unrestricted?
    • How can you manage confidential material in a safe way?
  • What are server hardware and setup needs?
    • How to utilize student laptops vs. workstations vs. large servers vs cloud?
    • Is cloud computing feasible?
    • Are there needs for special hardware (GPUs, many-core, large memory systems)?
  • Education
    • What tools and features are missing?
    • What design models, PDKs, IP are needed for classes?
    • What process nodes are adequate?
    • Are there different needs for fabrication vs. simulation only?
    • What are the deficiencies of open-source hardware and EDA tools?
  • Research
    • What advanced process nodes are needed?
    • What level of technology details is adequate?
    • What manufacturing information is needed?
    • Do you need partially processed CMOS wafers for further in-house processing?
    • Do you need wafer information for yield/analysis?
    • How to collaborate and fund fabrication?
    • What is needed for post-Moore technologies?

For all:

Diversity and Inclusion:

  • How can we increase hiring pools?
  • How can we build a community college pipeline?
  • How can we increase undergraduate enrollments?
  • How can we increase graduate enrollments?
  • How can we broaden recruitment to improve diversity?
  • How can the community be made more inclusive?

Existing resources:How are the following adequate or inadequate? What must be changed to make them adequate?

  • MPW shuttles
  • Cost of fabrication
  • Predictive technologies and PDKs: (e.g. FreePDK, ASAP, PTM)
  • Open-source, manufacturable PDKs (e.g. Skywater130)
  • Availability of design collateral for research and fabrication
  • Open-source designs, IP, and EDA tools
  • ISAs (RISC-V, MIPS, etc.)
  • Digital IP
  • Analog IP
  • EDA tool flows (OpenRoad, OpenLane, gEDA, etc.)
  • Other?

Areas of flexibility: What areas can we make compromises?

  • Can fabrication companies make models and PDKs available to a broad set of universities under NDA?
  • Can EDA companies imagine a way to have their tools on a centralized server, reduce needs for NDAs, or otherwise reduce the cost of ownership?
  • Can open-source tools reduce the cost and difficulty of maintenance of commercial tools while providing the same research and educational benefits?
  • What design collateral can be made open/freely available to participating universities?
    • Cell libraries, memory generators, PLLs, processors, peripherals?
    • Chiplets (“fill in the blank” design)?
  • Can open-source IP be used instead of commercial IP?
  • Access to fabrication for universities:
    • What technologies are sufficient?
    • How should the available technologies be updated over time?
    • How should it be funded and managed?
  • How can recruiting pathways from universities to companies and startups be facilitated?
  • Can the research and education infrastructure be extended to help startups?
  • Other?