Low-cost terahertz SoC takes shape as Taiwan advances next-generation communications

Taiwan, China's push toward low-cost terahertz SoC is a timing bet: achieve monolithic integration before 6G standards freeze (2028-2030), displacing discrete D-band/W-band front-end modules. The technical cascade is direct—InP and GaN epitaxy yield targets converge toward CMOS benchmarks, EDA mm-wave simulation modules gain outsized weight, and packaging migrates from FC-BGA to 3D stacking. The real compliance risk sits upstream in equipment, not in the chip itself. Export controls on advanced lithography and packaging tools create a structural 'can design but struggle to mass-produce' bottleneck; the low-cost strategy implicitly concedes process-node compromises—a soft underbelly competitors will exploit. On market dynamics, Qualcomm and Samsung have three years of W-band 6G prototyping; Nokia holds D-band academic depth. If SoC-level cost erosion materializes, the first casualty isn't infrastructure vendors but Qorvo and Skyworks' valuation thesis—their moat rests on module-level integration, and SoC consolidation undercuts that entirely. The 12-24 month long tail: stop fixating on 6G commercialization. The incremental battleground is THz spillover into 4D automotive radar and industrial NDT. Once SoC unit cost drops below $50, non-communications adoption will outpace every current forecast.