Sat 15 Aug 2026 / 11:55 ET
Kernel
Hardware 4 min read

Silicon wafer sizes, from 1 inch to the stalled 450 mm leap

Wafer size means diameter: a guide to inch names, 200 mm and 300 mm production, and why area is not chip yield.

Felix Aranda

By Felix Aranda / Silicon Editor

Silicon wafer sizes refer to a wafer’s diameter, not its thickness. Common production names include 150 mm, 200 mm and 300 mm, usually called 6-, 8- and 12-inch wafers. Those inch labels are inherited conventions, not exact conversions. SEMI says semiconductor silicon wafers are produced in diameters from one inch to 12 inches. A 450 mm, or nominal 18-inch, wafer remains developmental in a 2025 supplier account, not a settled production standard.

A wafer is the thin, round silicon substrate on which manufacturers build circuits through deposition, etching and photolithography, then dice the finished circuits into individual dies. Diameter determines how much circular surface one processing pass can cover. It does not determine how many working chips emerge.

Silicon wafer sizes: the diameter chart

  • 25.4 mm: 1 inch
  • 51 mm: 2 inch
  • 76 mm: 3 inch
  • 100 mm: 4 inch
  • 125 mm or 130 mm: nominal 5 inch
  • 150 mm: nominal 6 inch
  • 200 mm: nominal 8 inch
  • 300 mm: nominal 12 inch
  • 450 mm: nominal 18 inch, developmental

Metric labeling became prevalent as wafer diameters grew, although the older inch names remain common. A 300 mm wafer, for example, measures about 11.8 inches across but is routinely called a 12-inch wafer. Tokyo Electron’s 2015 historical account traces the industry’s progression through 2-, 3-, 4-, 5-, 6- and 8-inch generations to 300 mm.

Diameter and thickness are different specifications

“300 mm” states a wafer’s width. It does not state its thickness, crystal orientation or processing specification. Thickness is a separate mechanical specification. One supplier lists reference thicknesses for silicon of 675 micrometers at 150 mm, 725 micrometers at 200 mm and 775 micrometers at 300 mm. These are examples rather than universal specifications; wafers made from other semiconductor materials can have different thicknesses at the same diameter.

Supplier guidance also distinguishes the edge markings used for orientation. Wafers below 200 mm use flats, straight ground sections at the edge, to indicate crystallographic orientation. At 200 mm and above, a small edge notch provides that indication.

Why 300 mm carries more potential chips than 200 mm

A wafer is a circle, so surface area rises with the square of diameter. Moving from 200 mm to 300 mm increases diameter by 50 percent and area by 125 percent. More surface lets a fabrication step process more potential die at once. The per-die benefit depends on yield, utilization, chip size and the unusable margin near the wafer edge.

Worked area comparison

These are bare circular-area figures, not usable-die counts.

  • 200 mm wafer area: 31,400 mm²
  • 300 mm wafer area: 70,650 mm²
  • Calculation: 70,650 ÷ 31,400 = 2.25
  • Result: a 300 mm wafer has 2.25 times the nominal circular area of a 200 mm wafer.

The same supplier figures put a 450 mm wafer at 159,000 mm², about five times the nominal area of a 200 mm wafer. That is not a promise of five times the usable die. Dies may not fit at the curved rim, designers reserve edge space, and defects can disqualify die. Larger wafers lower cost per die only when a fab maintains yield and utilization.

200 mm vs. 300 mm vs. 450 mm

  • 200 mm (8 inch): 31,400 mm² of nominal area. A wafer supplier’s 2025 account identifies 200 mm and 300 mm as the most common sizes in use.
  • 300 mm (12 inch): 70,650 mm², or 2.25 times the area of 200 mm. In 2015, Tokyo Electron described 300 mm as the contemporary production era for logic, memory and analog lines.
  • 450 mm (18 inch): 159,000 mm², roughly five times the area of 200 mm. The 2025 supplier account characterized it as largely R&D and pilot work, citing the cost and infrastructure demands of adoption.

Changing wafer diameter requires compatible equipment. Supplier accounts describe new or adapted tools, automation and cleanroom changes as part of a move to larger wafers. The area calculation is easy. Reworking a production line is not.

What the wafer number tells you, and what it does not

  • It tells you: the nominal diameter, the equipment family it must fit, and the starting circular surface area for patterned circuits.
  • It does not tell you: exact usable-die count, yield, chip cost, thickness, material quality, crystal orientation or whether a particular fab can run it.

Keep the scope straight when reading shipment figures. SEMI’s quarterly silicon-wafer series covers semiconductor applications and includes polished, epitaxial and non-polished wafers shipped to end users. It excludes solar applications.

Frequently asked questions

What is the difference between a 200 mm and a 300 mm silicon wafer?

They are 200 mm and 300 mm in diameter, conventionally called 8-inch and 12-inch wafers. A 300 mm wafer has 70,650 mm² of nominal circular area, versus 31,400 mm² for a 200 mm wafer, or 2.25 times as much. That does not ensure 2.25 times as many working chips because chip size, edge losses and yield affect the result.

Why are silicon wafer sizes referred to as both inches and millimeters?

Inch names reflect earlier industry practice. Tokyo Electron’s 2015 historical account says metric terminology became prevalent around and after the 200 mm generation, while inch labels persisted as conventions. Thus 150 mm is commonly called 6 inch, 200 mm 8 inch and 300 mm 12 inch.

Are 450 mm silicon wafers used in commercial semiconductor production?

The supplied evidence does not establish broad commercial production. A wafer supplier’s 2025 account described 450 mm wafers as largely in R&D and pilot phases and cited cost and infrastructure demands as barriers to mass adoption. Claims about future adoption remain time-sensitive.

Sources

More Hardware/

view all ↗