Slide 1: Intro to Newsvendor Model and Applications In Semiconductors
Intro to Newsvendor Model and Applications In Semiconductors Reviewed 2025 © 2020-2025 Micron Technology, Inc. All rights reserved. Information, products, and/or specifications are subject to change without notice. All information is provided on an “AS IS” basis without warranties of any kind. State
Slide 2: Copyright guidelines
Copyright guidelines By using any content provided by the Micron Educator Hub, you acknowledge that Micron Technology, Inc. (“Micron”) is the sole owner of the content and agree that any use of the content provided by the Micron Educator Hub must comply with applicable laws and require strict compli
Slide 3: How to cite sources from the Micron Educator Hub
How to cite sources from the Micron Educator Hub • Micron is committed to collaborate with Use case How to cite sources educators to make semiconductor memory Whole slide deck or whole No additional citation required education resources available through the document Micron Educator Hub Description:
Slide 4: Table of Contents
Table of Contents 1. Goal, Objectives, and Target Audience 2. Intro to Newsvendor Problem 3. Using Statistics to Identify Solution 4. Apply Newsvendor Model in Micron Supply Chain 4
Slide 5: Newsvendor Model – goal and objectives
Newsvendor Model – goal and objectives Goal Objectives • Participants will be able to understand basic • Understand the Newsvendor model as a fundamental concepts of the Newsvendor Model and how it theory in supply chain management applies to the manufacturing industry at large, and • Explain how th
Slide 6: Target Audience
Target Audience • Individuals interested in a career in supply chain and those interested in Pro tip how data driven supply chain works Everyone interviewing at Micron can • Interns, NCGs (New College Grads), and new employees in supply chain use this presentation to prepare for roles need to unders
Slide 7: Intro to Newsvendor Problem
Intro to Newsvendor Problem 7
Slide 8: The Problem for a Newsvendor
The Problem for a Newsvendor John decides to open a Sets / Identifies the following Newspaper stall. parameters: ▪ Price (p = $ 1.00) ▪ Cost (c = $ 0.40) ▪ Scrap Value (S = $ 0.10) 8
Slide 9: The Inventory Decision
Newsvendor Model The Inventory Decision Inventory/Supply Uncertain Demand 60 people 60 people # of unsatisfied customer = 10 50 units Loss of profit = $6.00 Per unit understocking cost = $0.60 9
Slide 10: The Inventory Decision
Newsvendor Model The Inventory Decision Inventory/Supply Uncertain Demand 40 people 40 people # of excess inventory = 20 60 units Loss of profit = $6.00 Per unit overstocking cost = $0.30 Micron Confidential 10
Slide 11: John’s Dilemmas
Newsvendor Model John’s Dilemmas John needs to decide how many inventories to stock before the 1 demand quantity is realized, knowing that unsold copies will be worthless at the end of the day. 2 Trade-off between the understocking cost of stocking too less versus the overstocking cost of buying too
Slide 12: A Hand’s On Approach to Find Optimum
Newsvendor Model A Hand’s On Approach to Find Optimum 1 John looks at historical data and sees that the demand of newspaper is: 50 𝑢𝑛𝑖𝑡𝑠 ± 20 𝑢𝑛𝑖𝑡𝑠 2 Demand lies between 30 𝑢𝑛𝑖𝑡𝑠 and 70 𝑢𝑛𝑖𝑡𝑠 3 Inventory decision depends on Overstocking Cost and Understocking Cost 4 If…
Slide 13: Pictorial Representation
Newsvendor Model Pictorial Representation Higher Over- Inventory Decision Higher Under- stocking cost stocking cost Min Demand Max Demand Since in John’s Case, Understocking Cost : Overstocking Cost = 2 : 1, John’s decision of inventory would be 2/3rd way of the demand horizon towards 70 units, that
Slide 14: John's Optimal Solution
Newsvendor Model Factors Affecting John’s Inventory Decision ▪ Overstocking and Understocking Costs Sensitivity of John's Solution w.r.t. ▪ Overstocking and Understocking Costs Overstocking Cost and Understocking Cost 70 ▪ Mean Demand (Central tendency of the 65 uncertain demand variable) John's Opt
Slide 15: John's Optimal Solution
Newsvendor Model Factors Affecting John’s Inventory Decision ▪ Overstocking and Understocking Costs Sensitivity of John's Solution w.r.t. ▪ Overstocking and Understocking Costs Mean Demand ▪ Mean Demand (Central tendency of the 140 uncertain demand variable) 120 John's Optimal Solution 100 ▪ Range o
Slide 16: John's Optimal Solution
Newsvendor Model Factors Affecting John’s Inventory Decision ▪ Overstocking and Understocking Costs Sensitivity of John's Solution w.r.t. ▪ Overstocking and Understocking Costs Dispersion of Demand ▪ Mean Demand (Central tendency of the 90 uncertain demand variable) 80 John's Optimal Solution 70 ▪ R
Slide 17: Using Statistics to Identify Solution
Using Statistics to Identify Solution 17
Slide 18: Using Statistics to Identify Solution
? Using Statistics to Identify Solution Capturing Demand Uncertainty ▪ The source of our problem is that the demand is uncertain. ▪ Uncertainty is generally captured by two measurements: ⎻ Measurement of central tendency ⎻ Measurement of dispersion ▪ Central Tendency reflects the average behavior of
Slide 19: A View of Demand Distribution
Using Statistics to Identify Solution A View of Demand Distribution Uniformly Distributed Demand 0.03 Probability Density 0.025 • Here the demand can be anything 0.02 Function 0.015 between 30 and 70 and each value is 0.01 equally likely to occur (The probabilities 0.005 are same between 30 and 70).
Slide 20: Standard Deviation or Sigma
Using Statistics to Identify Solution Standard Deviation or Sigma Scatter Plot with Mean 50 and SD 11.25 100 ▪ Dispersion is a significant factor affecting the 80 Observation optimal inventory level. 60 40 ▪ A crude measurement of dispersion is range, as 20 used by John. 0 0 20 40 60 80 100 ▪ A more
Slide 21: Quality of John’s Solution
Using Statistics to Identify Solution Quality of John’s Solution ▪ John’s approximation is not bad. John’s Expected Profit Curve ▪ For uniformly distributed demand, John’s Expected Expected Profit ProfitCurve Curve approximation gives exact solution. $30.00 Understocking Cost 2 $25.00 Overstocking C
Slide 22: The Cost Diagrams for the Newsvendor
Using Statistics to Identify Solution The Cost Diagrams for the Newsvendor Expected Trade-off Profile (Units) Expected Trade-off Profile (Cost) 25 $14.00 Expected Number of Units $12.00 20 $10.00 Expected Costs 15 $8.00 10 $6.00 $4.00 5 $2.00 0 $- 30 35 40 45 50 55 60 65 70 30 35 40 45 50 55 60 65 7
Slide 23: Using Statistics to Identify Solution
Using Statistics to Identify Solution Till now, we talked about optimizing inventory level. A more general approach is optimizing service level. (Independent of the demand distribution). High / Low Service level means happy / unhappy customers. If we are meeting the demand, then our customers are ha
Slide 24: The Trade-Off
Using Statistics to Identify Solution The Trade-Off Benefits Costs We need to optimize the Low Stock-outs, High Inventory, High Service Level Happier Obsolesce Customers Service Level Low Inventory, High Stock-outs, Less chance of Unhappy obsolesce Customers Low 24
Slide 25: Required Parameters
Using Statistics to Identify Solution Required Parameters Demand Distribution ? • Normal • Mean = 50 • SD = 10 ▪ Price (p = $ 1.00) ▪ Cost (c = $ 0.40) ▪ Scrap Value (S = $ 0.10) 25
Slide 26: Optimization: Blunt Force Method
Using Statistics to Identify Solution Optimization: Blunt Force Method 1 Expected Profit Curve Take different “Inventory” values. $30.00 $25.00 2 Expected Total Profit Calculate the Expected Total Profit for all “Inventory” values. $20.00 3 $15.00 Plot the Expected Total Profit on a graph. $10.00 4
Slide 27: The Cost Diagrams
Using Statistics to Identify Solution The Cost Diagrams Expected Trade off Profile (Units) Expected Cost Profiles 25 14 12 20 Expected No. of Units 10 Expected Cost 15 8 6 10 4 5 2 0 0 30 35 40 45 50 55 60 65 70 30 35 40 45 50 55 60 65 70 Inventory Inventory Expected Overstocking # Expected Understo
Slide 28: Law of Diminishing Marginal Utility
Using Statistics to Identify Solution Law of Diminishing Marginal Utility • Also known as “Gossen's First Law” in Economics. • The marginal utility of each homogenous unit decreases as the supply of units increases. Total Utility Marginal Utility 25 12 10 20 8 Total Utility Marginal Utility 15 6 10
Slide 29: Newsvendor’s Marginal Utility Curve
Using Statistics to Identify Solution Newsvendor’s Marginal Utility Curve • For the newsvendor, total utility is equivalent to total profit, and marginal utility is equivalent to marginal contribution, i.e., marginal revenue – marginal cost. Total Utility Curve Marginal Utility Curve $30.00 $0.70 Ex
Slide 30: Marginal Utility Calculation
Using Statistics to Identify Solution Marginal Utility Calculation BENEFIT COST ▪ By ordering an additional unit, we are saving ▪ By ordering an additional unit, we are incurring the firm from stockout for one additional the cost of holding that additional unit, termed customer, termed as understock
Slide 31: Optimum Order Quality
Using Statistics to Identify Solution Optimum Order Quality ▪ At optimum (𝑄 = 𝑄 ∗ ), marginal utility of an additional unit = 0. ▪ Therefore, 𝐶𝑢 × Prob Demand > 𝑄 ∗ − 𝐶𝑜 × Prob Demand ≤ 𝑄 ∗ = 0 ▪ But demand follows normal distribution with mean 𝜇 and standard deviation 𝜎. Prob Demand >…
Slide 32: Graphical Representation of the Solution
Using Statistics to Identify Solution Graphical Representation of the Solution • Service level: 𝛼 = Prob Demand ≤ Inventory on hand 𝐹 𝑄 is the Cumulative PDF of the Actual Demand Distribution Function, Area under the therefore synonymous to 0.045 curve = Optimal Service Level 0.04 Probability…
Slide 33: Optimum Service Level for John
Using Statistics to Identify Solution Optimum Service Level for John 𝐶𝑢 = 𝑝 − 𝑐 = $1.00 − $0.40 = $0.60 𝐶𝑜 = $0.40 − $0.10 = $0.30 0.6 𝐹 𝑄∗ = = 67% 0.6 + 0.3 ▪ Assuming Demand follows normal distribution with mean 50 and standard deviation 10: 𝑄 ∗ = 𝑁𝑂𝑅𝑀. 𝐼𝑁𝑉 67%, 50,10 ≈ 54…
Slide 34: Graphical Representation of Uniform Demand Solution
Using Statistics to Identify Solution Graphical Representation of Uniform Demand Solution • Service level: 𝛼 = Prob Demand ≤ Inventory on hand Uniformly Distributed Demand Scenario Area under the 0.03 curve = Optimal Probability Density Function Service Level = 0.025 𝐶𝑢 𝐶𝑜 +𝐶𝑢 0.02 0.015…
Slide 35: Application examples
Application examples 35
Slide 36: Newsvendor Model in Real Business Scenario
Newsvendor Model in Real Business Scenario Applying Newsvendor in Supply Chain Understocking Cost Overstocking Cost Demand Uncertainty Inventory Safety Inventory Depreciation Stock and Excess Cost Flexibility in Missed Opportunity Depreciation Cost of Capacity Decision (Product Margin) Idle Capacity
Slide 37: Newsvendor Model in Semiconductors
Newsvendor Model in Semiconductors Other Applications of Service Level 1 Trade off between inventory cost and Production Quantity / Site Utilization underutilization cost + cost of missed Decisions opportunity Trade of between And many 2 Capacity Related • Site Qualification Decisions • Outsourcing
Slide 38: © 2020-2025 Micron Technology, Inc. All rights reserved. Information, products, and/or specifications are subject
© 2020-2025 Micron Technology, Inc. All rights reserved. Information, products, and/or specifications are subject to change without notice. All information is provided on an “AS IS” basis without warranties of any kind. Statements regarding products, including statements regarding product features,