Chip Binning
Chip binning is why two processors that came off the exact same design and the exact same wafer can be sold under different names, at different speeds, and at very different prices: manufacturing variation means not every die performs identically, so a fab sorts finished chips into performance "bins" after fabrication rather than throwing away everything that falls short of the top spec.
Manufacturing variation is unavoidable even on an otherwise perfect wafer, because no fabrication process is perfectly uniform across a 300-millimeter disc. Tiny differences in transistor threshold voltage, dopant concentration, and dimensional control from the center of a wafer to its edge mean two dies built from an identical layout still behave slightly differently — one may tolerate a higher clock speed before errors appear, another may leak more current at the same voltage, and another may have one defective core out of several identical ones on the same die.
Binning turns manufacturing variation into a business advantage instead of pure waste. Rather than scrapping every die that cannot hit the flagship product's clock speed, a company sells it as a lower-tier product at a lower price — a chip with one defective core out of eight can have that core permanently disabled and be sold as a six- or seven-core part, and a chip that cannot sustain the highest clock speed can be sold as a lower-clocked, lower-priced model. This is also why some enthusiasts can "overclock" a lower-tier chip closer to flagship performance: the physical silicon may be nearly identical, and the bin assignment was sometimes conservative rather than a hard physical limit.
Binning strategy is a real lever in a company's product-line economics, not just a testing formality. When yields are excellent and most dies qualify for the top bin, a manufacturer can end up with more "flagship-grade" chips than the premium market can absorb, so some get deliberately sold into lower tiers even though they could have passed at the top spec — demand forecasting, not just physical capability, shapes which bin a given chip ends up in.
| What testing reveals | Typical outcome | Business result |
|---|---|---|
| Meets top clock speed and full core count | Premium bin | Sold as the flagship SKU |
| Meets reduced clock speed or has disabled cores | Mid-tier bin | Sold as a cheaper, lower-spec SKU |
| Fails minimum functional spec | Reject | Scrapped, or salvaged for a simpler product |
Binning is also a direct byproduct of yield economics upstream in the fab. A mature, well-controlled process node produces a narrower spread of die performance and fewer rejects, which pushes more dies into the top bin; a newer, less mature node typically shows wider variation, so early production runs on a brand-new node often see a smaller fraction of dies qualifying for the flagship bin until the process itself tightens up over time.
Read binning through a yield-economics lens rather than a "good chip, bad chip" lens: nearly every die on a wafer started from the identical design and mask set, and the price tag it eventually carries reflects how that specific physical die performed on the test floor — not a difference in what was designed, but a difference in what one small patch of silicon turned out to tolerate.