Dynamically determining yield expectation
View Patent ↗A yield expectation determination is dynamically provided during manufacturing of a lot of integrated circuits. In one embodiment, the determination includes initially establishing a yield expectation for a lot, which can be based on a process grade, and adjusting the yield expectation during manufacturing based on test data from the kerf of a wafer. In addition, the yield expectation can be adjusted based on inspection data from optical and SEM inspection tools during manufacturing. Correlation coefficient models that correlate kerf data and inspection data to a yield expectation adjustment are used to dynamically adjust the yield expectation, resulting in a more accurate yield projection during manufacturing. The correlation coefficient models and/or process grade estimates are updated based on actual yield from a previous lot, thus further improving yield expectation accuracy.
1. A method for determining a yield expectation in an integrated circuit manufacturing facility, the method comprising the steps of:
initially establishing a yield expectation for a lot of wafers; and
adjusting the yield expectation during manufacturing of the lot based on test data from a kerf of a wafer.
2. The method of claim 1 , further comprising the step of establishing a process grade scale, and wherein the initially establishing step includes establishing the yield expectation for the lot based on a process grade from the process grade scale, each process grade including an initial yield expectation.
3. The method of claim 2 , further comprising the step of updating the process grade scale based on an actual yield of the lot.
4. The method of claim 1 , wherein the adjusting step further includes adjusting the yield expectation during manufacturing based on inspection data from at least one inspection tool.
5. The method of claim 4 , wherein the adjusting step further includes adjusting the yield expectation based on at least one correlation coefficient model that correlates at least one of the following: the inspection data to a yield expectation adjustment, electrical test data from the kerf to a yield expectation adjustment, and physical test data from the kerf to a yield expectation adjustment.
6. The method of claim 5 , wherein the at least one correlation coefficient model includes a correlation coefficient model for at least one of the following kerf tests: a front-end-of-line (FEOL) critical dimension measurement, a back-end-of-line (BEOL) ring oscillator performance/power test, a BEOL electrical defect test, a BEOL memory array test, a BEOL antennae macro test and a BEOL health-of-line test.
7. The method of claim 4 , further comprising the step of updating at least one correlation coefficient model based on an actual yield of the lot.
8. The method of claim 1 , wherein the adjusting step includes adjusting the yield expectation based on at least one correlation coefficient model that correlates at least one of the following: inspection data to a yield expectation adjustment, electrical test data from the kerf to a yield expectation adjustment, and physical test data from the kerf to a yield expectation adjustment.
9. The method of claim 8 , further comprising the step of updating each correlation coefficient model based on an actual yield of the lot.
10. A system for determining a yield expectation in an integrated circuit (IC) manufacturing facility, the system comprising:
an establisher for initially establishing a yield expectation for a lot of wafers; and
an adjuster for adjusting the yield expectation during manufacturing of the lot based on test data from a kerf of a wafer.
11. The system of claim 10 , wherein the establisher further establishes a process grade scale, and establishes the yield expectation for the lot based on a process grade from the process grade scale, each process grade including an initial yield expectation.
12. The system of claim 11 , further comprising an updater for updating the process grade scale based on an actual yield of the lot.
13. The system of claim 10 , wherein the adjuster further adjusts the yield expectation during manufacturing based on inspection data from at least one inspection tool.
14. The system of claim 13 , wherein the adjuster further adjusts the yield expectation based on at least one correlation coefficient model that correlates at least one of the following: the inspection data to a yield expectation adjustment, electrical test data from the kerf to a yield expectation adjustment, and physical test data from the kerf to a yield expectation adjustment.
15. The system of claim 14 , wherein the at least one correlation coefficient model includes a correlation coefficient model for at least one of the following kerf tests: a front-end-of-line (FEOL) critical dimension measurement, a back-end-of-line (BEOL) ring oscillator performance/power test, a BEOL electrical defect test, a BEOL memory array test, a BEOL antennae macro test and a BEOL health-of-line test.
16. The system of claim 13 , further comprising an updater for updating at least one correlation coefficient model based on an actual yield of the lot.
17. The system of claim 10 , wherein the adjuster adjusts the yield expectation based on at least one correlation coefficient model that correlates at least one of the following:
inspection data to a yield expectation adjustment, electrical test data from the kerf to a yield expectation adjustment, and physical test data from the kerf to a yield expectation adjustment.
18. The system of claim 16 , further comprising an updater for updating each correlation coefficient model based on an actual yield of the lot.
19. A program product stored on a computer-readable medium, which when executed, determines a yield expectation in an integrated circuit (IC) manufacturing facility, the program product comprising:
program code for initially establishing a yield expectation for a lot of wafers; and
program code for adjusting the yield expectation during manufacturing of the lot based on test data from a kerf of a wafer.
20. The program product of claim 19 , wherein the establishing code establishes at least one correlation coefficient model that correlates the test data from the kerf to a yield expectation adjustment, and the adjusting code adjusts the yield expectation based on the at least one correlation coefficient model.