At a Glance
Objective
- Students will read and interpret real historical price data (a PC price index and smartphone average selling price) to explain a supply-and-demand trend.
- Students will compare and contrast consumer-grade and AI data center-grade hardware, including price, purpose, and manufacturing complexity.
- Students will identify which regions lead each stage of semiconductor production and explain why that concentration creates supply-chain risk.
- Students will explain how each major system inside an AI data center maps to an equivalent part of a personal computer.
- Topics Covered
- computing systemssupply and demanddata literacyAI
- Subjects
- Computer Science
- Related Courses
- Intro to Software TechIntro to Hardware Tech
- Type of Lesson
- discussiondata-analysis
Related Unit(s)
Lesson Plan Notes Why Did RAM Get So Expensive? (A High School Lesson on AI and Computer Part Prices)
I didn’t introduce the international impacts of AI until much later in the school year, during a unit on Impacts of Technology, but here in the hardware unit is the first place it naturally fits. In my class, a handful of students built computers (usually for gaming), and those students recognized that GPUs are expensive. Connecting these price shifts to real-life situations makes the lesson more compelling and helps students see how we are all affected by the growing demand for AI, data centers, and the raw materials semiconductors require. Because this is the most abstract lesson in the unit, I keep it relevant with real-world data and questions about the data.
What this lesson actually is
- Why did RAM get so expensive? (page 1): the hook, a guess, then real historical PC and smartphone price data.
- Projected 2026 price impact (page 2): a moderate vs. pessimistic scenario table plus four analysis questions that work with any year’s data.
- Gaming vs. AI data center hardware (page 3): a Venn diagram comparing a consumer graphics card with AI data center hardware on price, purpose, and physical complexity, plus a short “So what?” prompt.
- Who makes what? (page 4): match each chip production stage (design, fabrication, memory, equipment, assembly/testing/packaging) to the region that leads it, then discuss what happens if one country’s factories shut down.
- Label the AI data center (page 5): match each data center system (compute, networking, memory, storage, power, cooling, orchestration software) to its PC equivalent.
- Take the tour (pages 6 and 7): an illustrated, numbered map of an AI data center with a tour-guide script, ending with a reflection on how similar (and different) it is to a student’s own computer.
This lesson will likely take more than one hour and run into 2 class periods. Day 1 covers the price story (pages 1 to 4); Day 2 covers the data center (pages 5 to 7). A “Full Circle” note on page 6 traces computers from room-sized (ENIAC) to pocket-sized (thanks to the integrated circuit) to building-sized again, because one data center now serves millions of devices. No slide deck or extra materials are needed. Every data point is cited on the worksheet itself.
Why real data (and a “scaled-up computer”) matters
Real data beats invented numbers. Students take it more seriously when the chart is real and cited, and they practice the habit of checking where a number came from. The lesson is designed to last. The graded skill is reading and reasoning from data, not memorizing 2026 prices. A teacher in 2028 swaps in a fresh chart; the questions still work. It fixes three common misconceptions head-on:
- “AI demand alone raised prices.” Actually, demand went up and supply for everyday devices went down, because factories shifted capacity to AI memory.
- “More advanced means smaller.” AI data center GPUs use physically larger, more complex packaging than gaming cards, which is part of why they’re harder to make.
- “The country that designs a chip also makes it.” Usually it doesn’t, and that separation is what makes this a supply chain.
The data center half makes AI feel less like magic. It uses the same parts students learned about in the components lesson, just far more of each. Right now the world is fixated on AI. We all want to know whether the investment will yield real value, so it’s important that students understand AI’s tangible foundation and the infrastructure required for it to function.
What I look for when I grade it
60 points total; the Exemplar has model answers for everything.
- Graph Reading & Analysis (15 pts): all 4 analysis questions answered with reasoning grounded in the actual charts, not guesswork.
- Venn Diagram (10 pts): at least 2 real similarities and 2 real differences, including price and complexity.
- Manufacturing Map (10 pts): each production stage matched to its leading region.
- Discussion & Prediction (10 pts): the prediction is reasoned from the data trend, not just asserted.
- Data Center Matching (10 pts): all 7 rows matched to their PC-part equivalent.
- Tour Reflection (5 pts): a genuine comparison between the data center and a personal device, not a restatement of the script.
The open-ended questions reward reasoning over “right answers.” A prediction that’s wrong but grounded in the chart can still earn full credit. This is also a formative assessment. It builds on students’ understanding of components and extends it to data centers and the comparable components required to power AI.
Where this goes next
- This is Lesson 5 of Unit 2 (Computing Systems). It zooms out from the device itself to the real-world market around it.
- Next up is Computing Systems & Costs: Building a PC, where students apply the same price awareness to a real constraint: a fixed budget and a build goal.