RF engineering is rebounding hard after decades of defense-only survival, driven by satellite constellations, 5G/6G rollout, and automotive radar simultaneously straining a thin talent pool.
Key Takeaways
Global spacecraft launches hit ~2,695 in 2024, up 10x from 2015; every satellite needs transceivers, antennas, filters, and amplifiers at each end.
5G MIMO radios integrate 64-256 transmit-receive chains vs. 2-4 in 4G, multiplying RF component demand 8-16x per base station across 642 operators.
IEEE survey data: 73% of EE employers cannot fill positions within six months, up from 45% five years ago; RF competes directly with the CHIPS Act semiconductor hiring frenzy for the same EE graduates.
Average RF engineer comp now exceeds $130K; top-end phased-array design roles list above $200K, with Mini-Circuits and Keysight funding university partnerships to bypass the broken academic pipeline.
The underlying physics (electromagnetics, thermodynamics, manufacturing tolerances) do not reduce to algorithms; link-budget fluency is learnable on the job, but phased-array design from scratch takes years.
Hacker News Comment Review
Commenters pushed back on the US-centric framing: the largest volume of RF engineering work has migrated to China, particularly for 5G and core switching, and catching up requires IP-origin legislation, not just manufacturing incentives.
Hardware practitioners noted the structural disadvantages that keep supply thin: lower salaries than software peers, mandatory on-site work, slow iteration cycles, sparse public documentation, and expensive tooling like HFSS and CST that has no mainstream open-source equivalent until very recently.
Hiring-side commenters confirmed the space-sector driver as dominant, with Amazon Kuiper and SpaceX creating second-order demand across LEO military programs; the signal is that demand is structural, not cyclical.
Notable Comments
@rhave: flags EMerge (open-source FDTD tool, released fall 2024) as a practical alternative to HFSS/CST for RF designers priced out of commercial licenses.
@jmarbach: Hubble Network made commodity Bluetooth LE chips talk to satellites; “antenna design and link budget work no software abstraction would fix.”
@commandlinefan: EE grads can pivot into software and back into RF; CS grads likely have the ability but will not get hired for RF roles regardless.
