As designed, the PRO COOL® process cost of ownership software does not consider rework loops. PRO COOL® is designed PRO COOL® cost of ownership software was designed to estimate the cost of ownership of a process sequence. This software assumes that the steps in the process sequence are contiguous or consecutive. In some situations, noncontiguous process sequences can be compared with PRO COOL® by replacing common intermediate steps with a dummy process step with zero cost of ownership. This application note outlines the limitations of modeling noncontiguous process sequences with PRO COOL® and describes a scenario where the comparison is successful and an unsuccessful scenario.

Assumptions and Limitations

PRO COOL® cost of ownership software assumes that the steps in the process sequence are contiguous or consecutive. This facilitates capacity analysis and allows summing the cost of ownership over a variable range of process steps. By making the following assumptions, process sequences can be compared by replacing common intermediate steps with a dummy process step with zero cost of ownership:

• Limited to comparing relative process COO of two or more process sequences.
• First and last steps in the process sequence must have specific TWO COOL® models.
• Bottleneck tools in each process sequence must have specific TWO COOL® models.
• Unique process steps each process sequence must have specific TWO COOL® models.

These limitations and assumptions will be illustrated in the following scenarios.

Successful Comparison

We compared the results of two PRO COOL® comparisons of the following process sequences. One comparison uses all process steps¹ and the other comparison replaces intermediate process steps with a dummy, zero cost of ownership process step:

Process A	Process B	Process X	Process Y
Wet Clean	Wet Clean	Wet Clean	Wet Clean
(200 hr MTBF)	(50 hr MTBF)	(200 hr MTBF)	(50 hr MTBF)
Coat Track	Coat Track
Expose	Expose	Dummy Step	Dummy Step
Develop Track	Develop Track
Defect Inspect	Defect Inspect	Defect Inspect	Defect Inspect

In all cases Defect Inspect is the bottleneck process step and limits the capacity of this process sequence. The difference in the comparisons is the mean time between failures (MTBF) of the two different Wet Clean process steps. All other process steps are identical between the process sequences.

In Processes X and Y, the dummy step has a COO of $0.00 and a very high throughput rate (10,000 wafers per hour) to insure that the dummy step is not a bottleneck. The dummy step has an inspection scrap of 0.03% to account for the difference in line yield by replacing three steps with one.

The following table summarizes the results from the management reports of PRO COOL® analysis on the four processes:

	Process A	Process B	A-B Delta	Process X	Process Y	X-Y Delta
Cost of Systems	$17,950K	$17,950K	0	$1,700K	$1,700K	0
# of Systems	4	4	0	3	3	0
Ave Utilization	51.1%	51.1%	0	39.48%	39.48%	0
Line Yield	99.95%	99.95%	0	99.95%	99.95%	0
Composite Yield	99.95%	99.95%	0	99.98%	99.98%	0
GWE	3722	3722	0	3722	3722	0
COO w/scrap	$37.5419	$37.5641	$-0.0222	$5.2445	$5.2667	$-0.0222
COO w/o scrap	$37.2919	$37.3141	$-0.0222	$5.1445	$5.1667	$-0.0222
Cost/Minute	$13.8648	$13.8730	$-0.0082	$1.9369	$1.9451	$-0.0082
Cost/Month	$605,613	$605,973	$-360	$84,602	$84,960	$-358
Cost/Month w/o Scrap	$601,581	$601,939	$-358	$82,989	$83,348	$-359
Fixed Costs	$13.2955	$13.2955	0	$1.2937	$1.2937	0
Time Based Costs	$5.3873	$5.3374	$-0.0499	$3.0975	$3.0475	$-0.0500
Wafer Based Costs	$18.8591	$18.9313	$-0.0722	$0.8533	$0.9255	$-0.0722

As this table clearly shows, replacing the intermediate steps with a dummy step changes the cost values but the difference between the options is virtually identical. Thus replacing intermediate steps with a dummy step may simplify the analysis without changing the comparative results. This is confirmed by the following cost driver comparison chart:

Click Here to View Enlarged Image

Unsuccessful Comparison

We repeated the comparison of the two processes with one change. We increased the number of Defect Inspection tools to two. In Processes A and B, the Coat/Develop Track is now the bottleneck. Since these process steps are replaced by a dummy step in Processes X and Y, the Wet Clean becomes the bottleneck for these processes. This changes violates the assumption that the bottleneck tools in each process sequence must have specific TWO COOL® models. The following table summarizes the results from the management reports of PRO COOL® analysis on the four processes:

	Process A	Process B	A-B Delta	Process X	Process Y	X-Y Delta
Cost of Systems	$19,150K	$19,150K	0	$2,900K	$2,900K	0
# of Systems	5	5	0	4	4	0
Ave Utilization	62.79%	60.53%	2.26%	45.26%	39.27%	5.99%
Line Yield	99.95%	99.95%	0	99.95%	99.95%	0
Composite Yield	99.95%	99.95%	0	99.98%	99.98%	0
GWE	5583	5382	201	6202	5382	820
COO w/scrap	$33.2217	$33.7960	$-0.5743	$5.2038	$5.9044	$-0.7006
COO w/o scrap	$32.9717	$33.5460	$-0.5743	$5.1038	$5.8044	$-0.7006
Cost/Minute	$18.4002	$18.0440	$-0.3562	$3.2018	$3.1524	$-0.0494
Cost/Month	$803,721	$788,163	$15,558	$139,854	$137,699	$2,155
Cost/Month w/o Scrap	$797,673	$782,333	$15,340	$137,166	$135,366	$1,800
Fixed Costs	$9.4685	$9.8223	$-0.3538	$1.3194	$1.5205	$-0.2011
Time Based Costs	$4.8940	$5.0424	$-0.1484	$3.0110	$3.4585	$-0.4475
Wafer Based Costs	$18.8591	$18.9313	$-0.0722	$0.8533	$0.9255	$-0.0722

As this table clearly shows, replacing the intermediate steps with a dummy step changes not only the cost values but also the differences between the options. The difference between this unsuccessful case and the previous case is the location of the bottleneck tools. The bottleneck tools in each process sequence must have specific TWO COOL® models.

References

D. L. Dance, "Cost of Capacity Mismatch ", Applied Co$t Modeling, Dec. 1996, Wright Williams & Kelly, Inc., Pleasanton, CA 94588.

Back to Top