A Complete Guide to Multi-Project Wafer Services for ASIC Teams
For IC design teams that need prototype silicon but cannot justify the cost of a full mask set, multi-project wafer (MPW) shuttle services have long been the standard answer.
An MPW shuttle combines designs from multiple customers onto a single wafer reticle or across multiple wafer lots, splitting the photomask cost among participants while each team receives its own packaged die.
This guide explains the mechanics of MPW services, helps you evaluate whether your program is a good fit, and outlines the planning steps that make the difference between a smooth shuttle experience and one filled with avoidable delays.
What Is an MPW Shuttle?
In a traditional full-mask tape-out, a single design occupies the entire reticle field. The customer pays for the complete set of photomasks, which at advanced nodes can run into hundreds of thousands of US dollars even before wafer fabrication begins.
An MPW shuttle changes this equation by placing multiple designs side by side within the same reticle area, or by aggregating designs across a fixed number of wafers. Each participant pays only a fraction of the total mask cost proportional to the area their design occupies.
The trade-off is that shuttle schedules are driven by the foundry's calendar rather than the individual customer's timeline. Participants must design to a common set of rules and align their tape-out readiness with the shuttle close date.
- Mask cost sharing is the primary economic benefit of MPW over full mask
- Shuttle schedules are typically quarterly or semi-annual depending on node demand
- Each customer receives diced and packaged parts as if they had done a standalone run
- IP licensing terms should be reviewed for MPW-specific restrictions
When Does MPW Make Sense?
MPW is ideally suited for programs where the primary goal is functional validation, characterization or early customer sampling before committing to volume production.
Typical use cases include university research chips, startup prototyping, IP validation vehicles and pre-production engineering samples for market testing.
Programs that already have firm volume commitments and tight schedule windows may find that the shuttle waiting time outweighs the mask-cost savings. In those cases, NTO (non-recurring tooling) or direct full-mask engagement may be more appropriate despite higher upfront cost.
- Good fit: First silicon for a new product category or technology validation
- Good fit: Low-to-moderate volume expectation (under ~10K units per year)
- Less suitable: High-volume products with aggressive time-to-market requirements
- Less suitable: Designs requiring custom metal stack or non-standard process options
Cost Comparison: MPW vs Full Mask
At mature nodes such as 180nm, 130nm or 90nm, a full mask set may cost between USD 30,000 and 80,000 depending on the number of layers and mask complexity. An MPW slot on the same node might reduce the per-team contribution to roughly USD 5,000–15,000 plus wafer share and packaging costs.
At 28nm or 14nm, the gap widens dramatically. Full mask sets can exceed USD 300,000–800,000, while MPW participation typically falls in the USD 20,000–60,000 range excluding IP and design services.
These figures vary significantly by foundry, region and shuttle organizer. Obtaining a detailed quote early in the planning cycle is essential for accurate budgeting.
Planning Checklist for Your First MPW Run
Success on an MPW shuttle depends less on the shuttle organizer and more on the preparation work done before the shuttle close date.
Below is a condensed checklist drawn from common failure modes observed across hundreds of shuttle participations.
- Confirm target foundry and process design kit (PDK) version availability at least 8 weeks before shuttle close
- Run DRC and LVS on your design using the shuttle-specific rule deck, not just generic foundry rules
- Verify IP license coverage includes MPW shuttle usage (some licenses restrict to full-mask only)
- Plan packaging option before tape-out — wirebond, QFN, QFP or BGA each have different pad ring and die size constraints
- Reserve test and evaluation board resources in parallel with shuttle submission
- Allow 4–8 weeks after wafer delivery for packaging, test and shipment depending on package type
How TOMPW Helps Global Teams Navigate MPW
For design teams located outside the primary semiconductor manufacturing regions, coordinating an MPW shuttle involves additional complexity around communication, payment logistics, shipping and customs clearance.
TOMPW Service has been helping ASIC teams navigate these challenges since 2014, providing a single point of contact for shuttle coordination, payment handling, packaging handoff and international logistics.
Whether this is your first MPW run or your twentieth, having a partner who understands both the technical requirements and the operational realities of cross-border semiconductor services can save months of trial-and-error learning.
