AI Prompt Library for AMD Stock Analysis

AMD has gained roughly 700% over the five years ending 2024, yet the stock regularly swings 10–15% on a single earnings print. That volatility is the opportunity — and the risk. The traders who consistently extract edge from AMD are the ones who show up to each catalyst with a structured research process, not a gut feeling.

Generic AI prompts produce generic output. Ask ChatGPT to ’analyze AMD’ and you get a Wikipedia summary. The real leverage comes from prompts engineered around AMD’s specific drivers: data center GPU attach rates, MI300X competitive benchmarks against H100, PC TAM cycles, and Lisa Su’s capital allocation signals. The difference in output quality is not marginal — it’s the difference between noise and a usable thesis.

This library gives you battle-tested prompts for every stage of the AMD research workflow — from pre-earnings positioning to post-print analysis to longer-term competitive mapping. Copy, paste, and adapt each prompt directly into your AI tool of choice.

Why AMD Demands Its Own Prompt Framework

AMD operates across four distinct segments — Data Center, Client, Gaming, and Embedded — each with its own cycle, margin profile, and competitive dynamic. A prompt built for a consumer discretionary stock will miss every relevant signal. You need prompts that pull on GPU market share progression, x86 server socket wins against Intel, and the Xilinx integration trajectory in embedded.

The stock also trades on expectations for MI300X and future Instinct GPU generations relative to NVIDIA’s roadmap. That means sentiment and forward guidance language carry outsized weight at earnings. A well-structured AI prompt can extract the specific language shifts between quarters that signal whether management is gaining or losing confidence in data center momentum.

Finally, AMD’s options market is deep enough that implied volatility before earnings often embeds a 10%+ expected move. Prompts that help you model that move against historical realized volatility can sharpen position sizing before you ever open a chart.

• Four reportable segments with non-correlated cycle timing
• MI300X GPU revenue is the single highest-scrutiny line item each quarter
• Server CPU share gains against Intel are tracked socket-by-socket by analysts
• Embedded segment (Xilinx) acts as a leading indicator for industrial capex cycles
• Options implied move at earnings routinely exceeds 10% — significant for sizing

Pre-Earnings Prompt: Build Your AMD Thesis in Minutes

The most valuable use of AI in AMD research is compressing the pre-earnings prep cycle. Instead of reading 15 analyst reports manually, one well-constructed prompt can surface the consensus expectation, the key debate, and the asymmetric scenario — in under two minutes.

The prompt below forces the model to separate what the market already knows (priced in) from what is genuinely uncertain. That distinction is where alpha lives. If every analyst is modeling $4.5B in data center revenue and AMD prints $4.7B, the beat only matters if the market wasn’t already positioned for upside. The prompt makes that logic explicit.

You are an equity research analyst specializing in semiconductors.
AMD reports earnings in [X days]. The current consensus estimate for Data Center revenue is [$X billion].
Task 1: Identify the 3 most contested variables analysts are debating heading into this print.
Task 2: For each variable, state what the bull case and bear case scenario implies for EPS.
Task 3: Based on AMD's last 8 earnings prints, describe the pattern in post-earnings price action relative to the size of data center revenue beats or misses.
Task 4: Flag any guidance language from the prior quarter that would have predicted this quarter's setup.

Competitive Positioning: AMD vs. NVIDIA Prompt

AMD’s data center GPU narrative is inseparable from NVIDIA’s. The MI300X is not evaluated in isolation — it is evaluated against H100 and B200 on performance-per-dollar, software ecosystem depth (ROCm vs. CUDA), and customer concentration. Any serious AMD thesis needs to stress-test the competitive gap explicitly.

The prompt below is structured to avoid the trap of recency bias. It asks the model to look at both the current benchmark gap and the trajectory — because a gap that is closing at 15% per generation tells a very different story than one that is stable or widening.

Use this prompt quarterly, updating it with the latest MLPerf benchmark data and hyperscaler capex commentary from AWS, Azure, and Google earnings calls.

Compare AMD MI300X against NVIDIA H100 and B200 across the following dimensions:
1. Price-per-FLOP at current street pricing
2. Memory bandwidth and capacity advantage/disadvantage for LLM inference workloads
3. ROCm vs. CUDA software ecosystem friction — quantify where possible
4. Hyperscaler public procurement signals (AWS, Azure, Google) from the last two earnings cycles
5. Conclude with a one-paragraph assessment: is AMD closing the competitive gap, holding steady, or losing ground — and what does that imply for 2025 data center revenue estimates?

Post-Earnings Debrief Prompt: Extract the Signal Fast

AMD earnings transcripts run 5,000–7,000 words. The most important content — guidance tone, segment-level inflection, and management hedging language — is distributed unevenly across that document. Reading it linearly is inefficient. A structured extraction prompt surfaces the highest-signal content in under 60 seconds.

Pay particular attention to how management discusses the Embedded segment. It was AMD’s fastest-growing division in 2022, then went into a prolonged inventory correction. The language around Embedded recovery has been a reliable leading indicator of the next leg of the stock’s broader cycle.

Run this prompt immediately after the transcript drops, before analyst notes hit your inbox. Forming an independent view before consensus recycles is how you avoid anchoring.

Below is the AMD Q[X] earnings call transcript. Perform the following analysis:
1. Extract every forward-looking statement about Data Center GPU revenue — flag any change in language intensity vs. last quarter.
2. Identify the exact moment Lisa Su addresses the Embedded segment recovery timeline — quote it directly.
3. List every instance where management used hedging language ('we expect,' 'we believe,' 'subject to') vs. declarative language ('we will,' 'we are on track').
4. Score overall guidance confidence on a 1–10 scale with a one-sentence rationale.
[Paste full transcript below]

AMD Options Strategy Prompt: Size the Expected Move

AMD’s 30-day implied volatility regularly spikes to 60–80% in the week before earnings, then collapses post-print regardless of direction. That IV crush is a known structural feature — and it punishes traders who buy options directionally without accounting for it. A prompt that models the expected move against historical realized outcomes reframes the sizing decision.

The prompt below does not make the trade for you. It gives you the inputs to make a defensible sizing decision: the market’s implied move, how often AMD has exceeded that move historically, and what the breakeven looks like for a straddle versus a directional spread.

• Pull AMD’s 1-day post-earnings realized moves for the last 8 quarters
• Compare each realized move to the options-implied move priced the day before earnings
• Calculate how often AMD exceeded the implied move to the upside vs. downside
• Model breakeven for an at-the-money straddle given current IV and premium
• Identify which strike has the highest open interest — this is where the market is hedged

Long-Term AMD Thesis Builder: 12-Month Framework

Short-term AMD trades live and die on quarterly catalysts. A 12-month position requires a different analytical frame: one that tracks the multi-year data center capex cycle, AMD’s wafer supply agreements with TSMC, and the PC market recovery arc. These variables move slowly but they determine the fundamental ceiling for EPS estimates two years out.

The prompt below constructs a bull/base/bear scenario model — not a price target, but a revenue and margin range — based on the key variables that will actually determine where AMD trades in 12 months. Use it to set your position thesis before entering and revisit it each quarter to check which scenario is materializing.

The most important discipline in using this prompt: write down your assumptions before you see the output. AI models can rationalize any scenario convincingly. Your edge comes from having a prior, then updating it — not from letting the model set your prior.

Build a 12-month bull/base/bear scenario model for AMD based on the following variable ranges:
- Data Center GPU revenue growth: bull +80% YoY / base +45% / bear +15%
- Client segment PC unit recovery: bull +12% / base +5% / bear flat
- Embedded segment return to growth: bull Q2 / base Q3 / bear Q4
- Gross margin trajectory: bull 54%+ / base 51-53% / bear below 50%
For each scenario: state the implied EPS range, the most likely stock price range at a 25x–35x forward P/E, and the single most important data point to watch that would move AMD from base to bull or base to bear.