Gallium-Arsenide RF Fab-Yield Leader — Aviation Radar

The mandate

The RF manufacturing director left after wafer-edge loss, ohmic-contact drift and passivation variation widened gain, noise and power distributions across radar MMIC lots. Product teams tune test limits by design, while the fab sees acceptable DC monitors that do not predict RF bins. The interim leader must restore physical-to-RF yield ownership without solving process instability through design-specific screening and expensive die selection.

Twelve months provide the recovery window. The first six weeks contain exposed lots, reconcile epitaxy through RF genealogy and establish product-level paretos. Months two through six close contact, gate, etch, passivation and wafer-uniformity mechanisms. Months seven through ten qualify equipment populations and stable RF test. The final two months install a permanent leader and transfer customer, supplier and release decisions.

Handover is achieved when functional yield, gain, noise figure, output power and reliability remain within approved distributions across qualified equipment and three epitaxy lots for ten weeks. The successor must chair one wafer disposition, approve a process-window change and defend residual RF margin to the radar customer. Engineering-only wafer regions and post-test die selection cannot count as recovered volume.

The leader may hold wafers, set process targets inside approved device rules, allocate engineering tools, approve RF test priorities, direct supplier containment and commit EUR 10 million within the authorised recovery envelope. They decide manufacturing release after reliability and product consultation. MMIC architecture, customer system requirements, export decisions, long-term epitaxy awards and final radar qualification remain with accountable executives.

Excluded are redesigning radar systems, changing aviation safety requirements, owning foundry-wide costs, negotiating export licences or replacing the entire RF process. The mandate covers GaAs physical yield, RF correlation and sustainable release for the selected MMIC families. Patent disputes, supplier liability and customer commercial remedies remain outside interim authority.

Why this seat is open

GaAs MMIC yield depends on the relationship among epitaxy, fine features, contacts, passivation, RF layout and measurement rather than DC monitors alone. The departure left fab and product teams optimising separate definitions of a good die. Temporary leadership can force physical mechanism closure and leave a successor with stable RF populations rather than selected wafer regions.

What you will own

• Reconcile substrate, epitaxy, wafer position, process equipment, mask, passivation, RF test and package genealogy.
• Connect DC monitors, physical dimensions and material evidence to gain, noise, power, efficiency and breakdown distributions.
• Direct mechanism splits across contacts, gates, etch profiles, surface treatment, dielectric and wafer-edge processing.
• Establish RF measurement-system control for calibration, fixtures, temperature, de-embedding, drift and cross-site correlation.
• Qualify tool groups and epitaxy lots using representative product corners instead of favourable monitor structures.
• Govern screening and bin limits around predictive value, escape consequence, false rejects and customer mission margin.
• Transfer process windows, open mechanisms, test standards, customer concessions and successor-led wafer releases.

Candidate qualifications

• Led GaAs or comparable RF compound-semiconductor yield recovery from wafer process through MMIC performance.
• Can evidence an RF loss mechanism invisible in otherwise passing DC process monitors.
• Understands epitaxy, ohmic and Schottky contacts, gate formation, passivation, MMIC test and RF reliability.
• Has controlled wafer holds and engineering-tool allocation during a constrained high-value product ramp.
• Directed product and fab teams through shared wafer maps, de-embedded measurements and physical analysis.
• Developed a successor able to reject die selection and test tuning as substitutes for process capability.

Non-negotiables

• Will lead onsite in Berlin and Ulm and attend Dresden and Toulouse technical councils.
• Has personally released volume RF compound-semiconductor wafers against product-level performance evidence.
• Accepts no authority over radar architecture, aviation requirements, export decisions or long-term supplier awards.
• Will disclose epitaxy, RF equipment, foundry, defence and aviation-customer relationships before appointment.

1. 49 words maximum. Which RF yield mechanism remained invisible in acceptable DC process monitors?
2. 49 words maximum. How would you prove wafer-edge loss is physical rather than an RF fixture artefact?
3. 49 words maximum. What must the successor demonstrate before inheriting MMIC wafer-release authority?