Video summary

A Fab Like This Doesn't Exist Anywhere Else in India Right Now

Main summary

Key takeaways

Technology

Technological concept & mission

  • HackerFab (HackerFab at IIT Bombay) aims to make semiconductor fabrication accessible to students in India, arguing that the industry is “highly gate-kept” because standard tools are extremely expensive.
  • The lab focuses on teaching and enabling fabrication of transistors (MOSFET-style) using a lower-cost fabrication “toolkit” whose principles still match industry processes.

Product features / capabilities described

Core machines/tools shown (for a basic MOSFET stack)

  1. Lithography machine

    • Creates patterns on wafers using a project-and-develop workflow with a photosensitive photoresist.

  2. Tube furnace (~1100°C)

    • Reaches up to ~1100°C with ~±2°C accuracy.
    • Used for oxide growth/annealing and for dopant activation, using a phosphorus-doped glass approach.

  3. Sputter system

    • Performs metal deposition via argon plasma sputtering under high vacuum.

Lithography approach & performance targets

  • The patterning concept is compared to a camera: photosensitive film → exposure → develop, where features shrink via optical projection.

  • The lab uses a projector-based lithography system:

    • a repurposed DLP projector
    • HDMI controls
    • a UV light source added
  • Current feature size: ~10 microns
    • Goal: improve to ~1 micron and further optimize, demonstrated via resolution test patterns.
  • Context is discussed using Moore’s law and typical industry nodes (e.g., 28 nm, 5 nm), but HackerFab emphasizes trading absolute node capability for cost/accessibility and training value.

Lithography cost-down strategy

  • The lab claims the lithography approach follows principles similar to high-end industrial systems (with ASML mentioned as an example), but implemented with a cheap projector.
  • Estimated cost:
    • currently about 2 to 2.5 lakhs
    • a goal of under 1 lakh by using smaller optics and potentially cheaper components.

Sputtering system engineering details

  • Vacuum pumping
    • Rotary vane pump down to ~0.05 mbar
    • then a turbo molecular pump for lower pressures
  • Vacuum measurement
    • Includes a Penning gauge reaching ~10^-5 mbar
    • base pressure reported around ~10^-6 mbar
  • Process mechanism
    • Argon is ionized into plasma.
    • Ions knock atoms off a metal target.
    • The atoms deposit onto the substrate.
  • Notes emphasize the need for tight pressure control, so plasma strikes efficiently without excessive collision in the gas phase.
  • Cooling uses a “hacky” but functioning CPU water cooling system for the sputter hardware.

Tube furnace building & reliability

  • Built from scratch using:
    • Quartz tube
    • Kanthal heating element
    • refractory cement
    • ceramic wool insulation
  • Control:
    • uses a control algorithm that cycles current based on thermocouple feedback
  • Reliability claim:
    • reported as highly reliable
    • built by two first-year students without prior experience, highlighting DIY capability.

Reviews / analysis style arguments

The speakers repeatedly frame HackerFab as:

  • A hands-on education layer missing from India’s broader semiconductor expansion.
  • A way to train students for process engineering and yield intuition, not just theoretical knowledge.
  • A potential catalyst for scaling to more colleges through open sourcing designs and procedures.

Guides / tutorials / documentation mentioned

Open-sourcing and community learning

  • The lab claims to open-source everything related to their process.
  • A Discord server supports community learning, where students can follow the workflow and ask for help.

Onboarding guidance for newcomers

  • Onboarding begins with a simple “spin coater” step to teach thin-film coating as a gateway before lithography.
  • A staged path is described:
    • build a simple device
    • build confidence
    • gather funding under ~1 lakh
    • build the cheapest lithography prototype

Scaling plan / future product vision

“Fab in a box” concept

  • Aims to package fabrication capability into a shipping-container-like system for colleges.
  • Claimed needs:
    • colleges/students only need power and water
    • raw materials/chemicals are supplied
  • Target pricing: around 30 lakhs

Outcome goal

  • For IIT Bombay: students should graduate having “taped out a transistor”—an end-to-end, applied fabrication outcome.

Timelines & team progress

  • They expect to build first transistors within about a month (from the interview time).
  • Timeline:
    • project started around August
    • by September, it reached a working proof of concept
    • conference feedback highlighted scaling needs and helped drive next steps.

Main speakers / sources

  • Caleb (interviewer/host; from Runtime)
  • Dev (HackerFab team member)
  • Jack (HackerFab team member)
  • Jay (walked through wafer cleaning + lithography)
  • Dave (leading lithography cost-reduction efforts)
  • Additional contributor referenced: Aryaman (discussed educational impact and outcomes)

Original video