TFLN Foundry Lead Times in China: MPW vs Dedicated Runs (2026)

Typical TFLN/LNOI foundry lead times in China in 2026 range from about five business days to two or three weeks for in-stock standard wafers, about six to twelve weeks for MPW shuttle access, twelve or more weeks for a dedicated run, and three to five months for 200mm MPW. These are indicative planning estimates only: China foundry capacity is expanding fast, queue positions move, and every timeline should be confirmed at RFQ.

Lead-time matrix

All times are indicative planning ranges, not supplier commitments. China TFLN foundry capacity is expanding quickly; confirm current queue, wafer availability, process scope, export review, packaging scope, and delivery terms during RFQ.

MPW vs dedicated run: which to choose

Choose MPW when the goal is design validation, process learning, or first optical/RF data and the design can tolerate shared shuttle rules. MPW usually compresses cash cost and calendar risk because masks, wafer starts, and process windows are coordinated across multiple customers. The trade-off is reduced control: shuttle dates, die area, process options, delivery formats, and metrology packages may be fixed.

Choose a dedicated run when the programme needs custom process splits, dedicated wafer allocation, packaging integration, unusually tight film or electrode requirements, or a production-quality qualification trail. Dedicated runs are slower and more expensive, but they are the right path when the buyer needs control over the mask set, wafer count, test structures, hold points, and failure-analysis loop.

Factors that shift lead time

• Crystal cut: X-cut, Z-cut, MgO doping, and non-standard orientation can move the wafer availability date.
• Film thickness and uniformity: custom LN thickness, BOX thickness, TTV, bow, warp, and trimming requirements add metrology and process time.
• Electrode process: travelling-wave electrodes, thick metals, plating, impedance targets, and RF probe-pad design can extend the foundry cycle.
• Packaging: bare die is faster than fibre-array coupling, RF package assembly, thermal control, burn-in, or evaluation-kit delivery.
• Mask and design readiness: DRC closure, reticle scheduling, test-structure completeness, and GDS revisions can dominate the apparent foundry wait.
• Export licensing: cross-border technical data, high-speed modulator performance, end use, destination, and MOFCOM dual-use review can add delay before fabrication or shipment.

How to use the estimates

Start by separating substrate procurement from device fabrication. If the project needs raw wafers, compare the current shortlist in the TFLN/LNOI supplier guide and qualify cut, thickness, TTV, BOX, bow, warp, and sample history. If the project needs a fabricated PIC or modulator, ask whether the path is an MPW shuttle, a dedicated run, a qualified chip purchase, or an integrated project.

For earlier-stage supplier mapping across photonics and advanced materials, see deep-tech sourcing. For supplier qualification, RFQ management, sampling, negotiation, and handoff, see the process.