28nm Wafer & MPW Cost Explained: Constraints, Risk, and When It Makes Sense

28nm is a turning point in ASIC development. It is often described as a “mature advanced node,” but in practice it represents the last node where many non-hyperscale teams can realistically operate. Beyond 28nm, ASIC economics, risk, and organizational requirements change dramatically.

This article explains what drives 28nm wafer and MPW cost, and why decisions at this node must be made with exceptional clarity.

Why teams consider 28nm

28nm is typically chosen when designs require:

• High logic density
• Strong performance per watt
• Advanced SoC-level integration
• Long product lifetime and volume potential

Common applications include:

• Networking and data-path ASICs
• High-performance embedded SoCs
• Multimedia, imaging, and acceleration devices
• Infrastructure and automotive platforms

At this node, the motivation is rarely “cost reduction” — it is capability enablement.

What drives 28nm wafer cost

At 28nm, wafer cost is driven by advanced-node realities, not process age:

• High mask count
• Very tight design rules
• Significant DFM and variability considerations
• Yield sensitivity to layout quality
• Expensive respins

While 28nm is still planar, it behaves economically like an advanced node. The cost of mistakes rises sharply.

28nm MPW: availability and intent

MPW exists at 28nm, but it is not a general-purpose prototyping platform.

Typical characteristics:

• Very limited shuttle availability
• Strict die size limits
• Minimal process option flexibility
• Strong emphasis on design readiness

28nm MPW is typically used for:

• controlled first silicon
• architecture confirmation
• limited learning before committing to production

It is not suitable for exploratory or unstable designs.

MPW vs full mask at 28nm

MPW can make sense at 28nm when:

• This is first silicon
• The design is already highly stable
• One learning cycle is sufficient
• Volume projections are still uncertain

Full mask is often the better choice when:

• Design verification is complete
• Volume expectations justify NRE
• Schedule predictability is critical
• Backend planning is mature

At 28nm, many teams move directly to full mask once feasibility is confirmed.

Backend, yield, and test dominate total cost

At 28nm, backend considerations often outweigh wafer cost:

• Advanced packaging
• High pin counts
• Complex test strategies
• Yield learning that favors production-like conditions

At this node, focusing on wafer price alone is misleading.

Schedule risk becomes strategic

MPW schedules at 28nm are:

• infrequent
• inflexible
• tightly controlled

Missing a window can add months to a project timeline, which may outweigh any MPW cost advantage.

For time-critical products, MPW must be evaluated as a risk-management tool, not a cost-saving one.

 Evaluate MPW vs full mask for 28nm

At 28nm, the question is not “Can we afford MPW?”

It is “Is MPW the right strategic step for this program?”

You can evaluate this based on:

• design maturity
• volume expectations
• schedule pressure
• organizational readiness

👉  Use the MPW vs Full Mask decision tool

Final takeaway

28nm is the last practical node for many ASIC teams before economics and risk fundamentally change.

MPW is still possible — but only when:

• design risk is already low
• backend planning is advanced
• schedules are realistic

At 28nm, ASIC decisions must be driven by strategy and discipline, not experimentation.