Can a Computer Science Student Be Taught to Design Hardware?

Original Article ↗ Hacker News Discussion ↗

Scope: Chip vs. PCB Design

  • Several commenters note the article is about silicon / chip design and verification, not PCB or device-level “hardware,” causing confusion in the thread.
  • Some share that much PCB work is copying vendor reference designs and routing; “hardware design” in chip companies usually means RTL, architecture, or verification, a very different niche.

Can CS Students Learn Hardware?

  • Many say yes: hardware and software share abstractions (state, concurrency, modularity); good software engineers can pick up digital design and especially verification.
  • Others push back, arguing that pushing PPA (power–performance–area) and dealing with timing, metastability, and microarchitecture requires deep, specialized knowledge and is not just “parallel programming.”
  • Cross-disciplinary people (HW/SW/architecture combined) are described as rare but extremely valuable.

Talent Shortage, Pay, and Mobility

  • Some doubt a real “talent shortage,” pointing to underemployed ECE grads and aggressive offshoring.
  • Views on pay conflict: in some regions and for cutting-edge chip work, salaries reportedly match or beat general software; others insist software reliably pays more, especially at big US firms, and see many engineers moving from hardware to software, not vice versa.
  • Hardware careers are seen as geographically concentrated, less flexible, and more easily outsourced when work is highly spec- and test-driven.

Tooling and Accessibility

  • Proprietary, expensive, fragile EDA tools (Cadence, Synopsys, vendor FPGA suites) are widely blamed for keeping hardware niche and discouraging experimentation.
  • Lack of open-source tools and open flows is cited as both cause and symptom of weak grassroots interest, though some newer open-source FPGA/ASIC initiatives are mentioned as promising but still limited.

Education and Curricula

  • Older CS programs often required substantial EE/architecture; many commenters report modern CS tracks dropping low-level courses, “deskilling” graduates for hardware roles.
  • Several advocate more hardware exposure for CS (FPGAs, HDLs, computer architecture) and more CS/software engineering for EE, but note departmental turf wars and lack of faculty interest.

Digital vs. Analog

  • Multiple commenters stress that analog/RF design is a very different, more physics-heavy discipline; CS backgrounds transfer poorly there, while digital logic/verification is more accessible to software people.