Fab Utilization Rate

Fab utilization rate is the percentage of a fab's total available wafer-processing capacity that is actually running production at a given time, and it matters more in semiconductor manufacturing than in almost any other industry because a fab's costs are overwhelmingly fixed. The equipment, cleanroom, and facility costs of a modern fab are committed the moment it is built, regardless of how many wafers subsequently run through it, so every wafer slot that sits idle is not a saved cost, it is a lost opportunity to spread that already-committed capital over one more unit of output, which is exactly why the capacity-reservation and take-or-pay mechanisms covered elsewhere in this cluster exist in the first place.

Fixed Costs Mean Idle Capacity Is a Direct Loss the same monthly cost is spread over however many wafers actually ran 90% utilization fixed cost spread over ~90% of capacity lowest cost per good wafer 55% utilization same fixed cost spread over ~55% of capacity far higher cost per good wafer Cost per wafer ≈ fixed facility cost ÷ wafers actually run Utilization is the denominator in that equation — it does not change the numerator at all. The empty grey slots in the right panel cost the same as a filled one — they just produce nothing.

This fixed-cost structure is also why fabs are far more reluctant than most manufacturers to cut prices during a downturn rather than simply idling capacity. Once a fab is built, running a wafer through it at a price barely above its marginal processing cost is still better than leaving that slot empty, since the fixed costs are being paid either way; a fab will frequently accept thin or even negative margins on marginal volume during a downturn specifically to keep utilization up, rather than protecting price at the cost of running fewer wafers through the same fixed cost base.

Fab's fixed costs are committed regardless of output
equipment, cleanroom, and facility costs do not scale down with fewer wafers
Cost per wafer depends almost entirely on utilization
same fixed cost divided by whatever volume actually ran
Two very different responses to a demand downturnthe fixed-cost structure favors one over the other
Protect price, let utilization fall
fixed costs spread over fewer wafers — cost per wafer rises
Cut price, keep utilization high
fixed costs spread over more wafers — cost per wafer holds
Fabs typically favor keeping utilization up
even at thin margins, over protecting price on half-empty capacity

This is the operational reality that the take-or-pay, capacity allocation, and customer concentration mechanisms covered elsewhere in this cluster all exist to manage upstream, since a fab that can lock in guaranteed volume through those contractual mechanisms avoids having to discover its own utilization shortfall in real time and react to it after the fact.

Utilization levelCost per waferTypical fab response
High (85%+)Near the lowest achievable for that fabProtect price, allocate scarce capacity carefully
Moderate (60-85%)Noticeably elevatedSeek incremental volume at reduced margin
Low (below 60%)Sharply elevated, often loss-making per waferAggressive price cuts, or capacity idling/closure decisions

Fab utilization is the operational metric that everything else in this cluster, from capacity-reservation contracts to the fab-scale terminology series, ultimately exists to protect — a gigafab's economics only work if its enormous fixed cost base is actually kept running near full capacity.

Take fab utilization rate further

Ask the copilot about this term, or have our engineers assess it against your process.