Digital Testing Lecture 3: Test Economics

Instructor: Shaahin Hessabi Department of Computer Engineering Sharif University of Technology

Adapted from lecture notes prepared by the book authors

Test Economics
Economics defined Costs Production Benefit - cost analysis Economics of design-for-testability Summary

Sharif University of Technology

Testability: Lecture 3

Page 2 of 17

Engineering Economics
Engineering Economics is the study of how engineers choose to optimize their designs and construction methods to produce objects and systems that will optimize their efficiency, and hence, the satisfaction of their clients.

Sharif University of Technology

Testability: Lecture 3

Page 3 of 17

Costs
Fixed cost: costs that dont change with use. Variable cost: increase with the production output. Total cost: sum of the above two costs. Average cost: total cost/number of units produced.
Example: Costs of running a car Fixed cost Variable cost Total cost Average cost
Sharif University of Technology

$25,000 20 cents/mile $25,000 + 0.2x 25,000 $ ----------- + 0.2

Purchase price of car Gasoline, maintenance, repairs For traveling x miles Total cost / x
Page 4 of 17

Testability: Lecture 3

Simple Cost Analysis

Case 1: 10,000 miles/yr, $12,500 resale value after 5 years 25,000 - 12,500 Average cost = $ ----------------------- + 0.2 = 45 cents/mile 50,000 Case 2: 10,000 miles/yr, $6,250 resale value after 10 years 25,000 - 6,250 Average cost = $ --------------------- + 0.2 = 38.75 cents/mile 100,000 Case 3: 10,000 miles/yr, $0 resale value after 20 years 25,000 - 0 Average cost = $ --------------- + 0.2 = 32.5 cents/mile 200,000

Sharif University of Technology

Testability: Lecture 3

Page 5 of 17

Cost Analysis Graph

40,000

Fixed, Total and Variable Costs ($)

25,000 20,000

Fixed cost

50

Va
0

ble ria

t os c

Average cost
ideal

50k

100k

150k

0 200k

Miles Driven
Sharif University of Technology Testability: Lecture 3 Page 6 of 17

Average Cost (cents)

tal To st co

100
(ag in re al g)

Production
Inputs (x): Labor, land, capital, enterprise, energy (x may include both fixed and variable costs) Production output: Q = f (x) Average product: Q / x Marginal product: dQ / dx

Sharif University of Technology

Testability: Lecture 3

Page 7 of 17

Law of Diminishing Returns

If one input of production is increased, keeping other inputs constant, then the output may increase, eventually reaching a point beyond which increasing the inputs will cause progressively less increase in output.
Output, Q(x)

Input Resources, x
Sharif University of Technology Testability: Lecture 3 Page 8 of 17

Technological Efficiency
Technological efficiency = Q/x where x = variable cost To maximize technological efficiency:
d Q ---- --- = 0; dx x 1.0 Tech. Eff. 0.5 dQ/dx 0.0 Input Resources, x dollars
Sharif University of Technology Testability: Lecture 3 Page 9 of 17

1 dQ Q -- ----- -- ---- = 0 x dx x2

or

Q dQ --- = ----x dx

Max. tech. eff. Q/x

Economic Efficiency
Maximum economic efficiency minimizes the total average cost X /Q, where X is the total (fixed + variable) cost. Maximum economic efficiency is achieved when total average cost equals the marginal cost, X /Q = dX /dQ. For average cost = marginal cost
Take variable cost to maximize technological efficiency Take total cost to maximize economic efficiency

Sharif University of Technology

Testability: Lecture 3

Page 10 of 17

Mass Production
Production can be increased at a faster rate than the increase of inputs. This is known as increasing returns to scale. Some reasons for increasing returns to scale:
Technological factors Specialization

If increase of inputs continues, eventually the law of diminishing returns applies.

Sharif University of Technology

Testability: Lecture 3

Page 11 of 17

Benefit-Cost Analysis
Benefits: Savings in manufacturing costs (capital and operational) and time, reduced wastage, automation, etc. Costs: Extra hardware, training of personnel, etc. Benefit/cost ratio:
Annual benefits B/C ratio = -------------------------- > 1 Annual costs
Sharif University of Technology Testability: Lecture 3 Page 12 of 17

Weighing Cost Against Benefit (example)

PCB average repair cost = $350, containing:
average cost of diagnostic test = $300 average price of a chip = $10 cost of replacing any number of chips + final go/no-go system test = $40

Shops repair log 95% of failed boards have a burnout failure of one of two specific chips
Sharif University of Technology Testability: Lecture 3 Page 13 of 17

Weighing Cost Against Benefit (example)

Test consultant replace those two chips without running any diagnostic test. Then:
cost of this repair = $60, since:
two chips cost $10 a piece, chip replacement and system test cost $40

expectation: only 5% of the boards will fail the system test.

full diagnostic test on each failed board and repairing it cost $350.

The average cost of repair = $(1 60 + 0.05 350) = $77.50 ( << $350)
Sharif University of Technology Testability: Lecture 3 Page 14 of 17

Economics of Design for Testability

Consider life-cycle cost; DFT on chip may impact the costs at board and system levels. Weigh costs against benefits
Cost examples: reduced yield due to area overhead, yield loss due to non-functional tests Benefit examples: Reduced ATE cost due to self-test, inexpensive alternatives to burn-in test

Sharif University of Technology

Testability: Lecture 3

Page 15 of 17

Benefits and Costs of DFT

Level Design and test +/+/+/Fabrication + + + Manuf. Test Maintenance test Diagnosis and repair Service interruption

Chips Boards System

+ Cost increase - Cost saving +/- Cost increase may balance cost reduction

Sharif University of Technology

Testability: Lecture 3

Page 16 of 17

Summary
Economics teaches us how to make the right trade-offs. It combines common sense, experience and mathematical methods. The overall benefit/cost ratio for design, test and manufacturing should be maximized; one should

select the most economic design over the cheapest design.

A DFT or test method should be selected to improve the product quality with minimal increase in cost due to area overhead and yield loss.
Sharif University of Technology Testability: Lecture 3 Page 17 of 17