Fabless RF power amplifier design for wireless, satellite, and defense
Falcomm designs RF power amplifiers and front-end modules for wireless, satellite, and defense applications. The engineering stack—MATLAB, Cadence, Synopsys, HFSS, PyTorch, TensorFlow—reflects a hybrid chip-design and ML-heavy operation. Active projects show early-stage ML integration into design workflows (model training/serving, ML model integration into chip design software) alongside core RF validation work, suggesting the team is automating parts of the design loop. Hiring is accelerating (18 of 26 roles posted in the last 30 days) and heavily weighted toward engineering and interns, consistent with a pre-production fabless startup scaling design and test capacity.
Falcomm is a fabless semiconductor company founded in 2021, headquartered in Atlanta, GA. The company designs power amplifiers and integrated RF front-end modules targeting wireless communication, satellite systems, and defense applications. The core intellectual property—a proprietary solid-state RF power amplification architecture developed at Georgia Tech—focuses on efficiency gains in a technology category that has seen minimal innovation over decades. The team operates across design, simulation, prototyping, and test, with active work on evaluation boards, automated test lab infrastructure, and material/component selection. Current operational challenges include thermal management, RF performance yield, volume production transition, and supply chain stability.
Falcomm uses MATLAB, Cadence, Synopsys, Mentor Graphics, HFSS, Agilent ADS, Keysight ADS, Spectre, Verilog, SystemVerilog, and Python/PyTorch/TensorFlow for RF circuit design, simulation, and ML-driven automation.
Falcomm is developing mm-wave power amplifier evaluation boards, automated test lab equipment, ML-integrated chip design workflows, 3D EM modeling tools, and transitioning prototypes toward volume production while addressing thermal and RF yield challenges.
Other companies in the same industry, closest in size