AMD and ARISTA NETWORKS Q3 Result Takeaways
High expectations meet execution reality, despite both companies growing strongly
The referenced earnings analysis reports were generated with the Earnings Analysis Report prompt from the professional prompt library on The INFERENTIAL INVESTOR.
Subscribe to access these tools and stock research.
Two more AI stocks reported Q3 earnings and are experiencing the same treatment by the market as Palantir.
AMD and ARISTA NETWORKS both beat on revenue and EPS in the quarter but its the expectations that count more after such strong stock price runs this year. Both are trading down after hours, continuing a theme of AI market darlings running into profit taking, that started with META and Microsoft. The market is showing that continually expanding valuation multiples were built on a foundation of accelerating growth. When this starts to slow, multiple start to derate rapidly as investors take profit.
We’ve summarized the highlights of each and linked the full reports for further detail.
ADVANCED MICRO DEVICES (AMD)
AMD just posted an outstanding Q3 2025, smashing analyst estimates with record revenue of $9.25 billion, a 36% jump year-over-year, and management declared it a “clear step up” in the company’s growth trajectory. This bullishness was validated by a “rapidly scaling” AI business, backed by massive new partnership announcements with OpenAI and Oracle. However, despite the strong beat and an above-consensus Q4 revenue guide of $9.6 billion, the stock puzzledly fell in after-hours trading. This appears to be a classic “sell the news” event; after a 100%+ run-up year-to-date, the stock was priced for absolute perfection, and the strong guidance likely wasn’t high enough to satisfy aggressive “whisper numbers,” giving profit-takers the perfect excuse to cash out.
ARISTA NETWORKS INC.
Arista Networks (ANET) reported a mixed Q3 2025 that led to a sharp after-hours sell-off, as a solid earnings beat was overshadowed by a weak forward-looking guide. While Arista surpassed Q3 consensus with $2.31 billion in revenue (+27.5% YoY) and $0.75 in adjusted EPS, its Q4 outlook signaled a clear normalization. The two main concerns driving the stock down were a significant deceleration in growth (with Q4 revenue guidance implying only 1.8% sequential growth) and, more critically, an unexpected and severe drop in Q4 non-GAAP gross margin guidance to 62-63%, a steep fall from 65.2% in Q3. This guidance suggests profitability is peaking, forcing investors to re-evaluate the company’s premium valuation as the era of easy beats and margin expansion appears to be ending.
Brief Follow up from the Arista Networks Earnings Call:
ANET Q3 2025: Flash Earnings Transcript Analysis
This was a classic case of a short-term “beat” being completely overshadowed by a forward-looking “guide-down.” The market’s negative reaction was not to the Q3 results themselves, which were strong, but to management’s Q4 guidance, which signaled a significant, near-term compression in profitability.
Here are the key themes from the call.
1. The Margin Story: The Single Biggest Driver of the Stock Drop
The entire negative story of the call revolves around the Q4 gross margin guidance.
The “Beat” (Q3): Arista reported a very high non-GAAP gross margin of 65.2%.
The “Guide-Down” (Q4): Management guided Q4 non-GAAP gross margin to 62-63%.
This implies a sequential drop of 220-320 basis points, a massive decline that immediately triggered alarms.
Management’s Explanation
On the call, management explained that the high 65.2% margin in Q3 was an anomaly, not a new baseline. A key snippet from the call stated that Q3’s strong margin was “influenced by favorable mix and inventory benefits“.
Translation:
“Favorable Mix”: In Q3, Arista likely sold a richer mix of high-margin enterprise products or software than expected, which offset the lower-margin, high-volume AI sales.
“Inventory Benefits”: This likely refers to selling through older, lower-cost inventory or other one-time supply chain accounting benefits that are now exhausted.
The Q4 guidance of 62-63% is not necessarily a “collapse” but rather what management considers a “normalization” back to the underlying business mix, which is increasingly weighted toward high-volume, “margin-dilutive” AI deployments. The market, however, saw this as a severe execution misstep or the beginning of a pricing war, and it punished the stock.
2. The Growth Story: Normalization is Spooking Investors
While revenue guidance was technically in line, the implied growth rate signaled a sharp slowdown.
Q3 Revenue Growth: +4.7% sequentially over Q2.
Q4 Revenue Guidance: The midpoint of $2.35B implies only +1.8% sequential growth over Q3.
CEO Jayshree Ullal’s new strategic narrative is “durable execution”. This is a pivot from the hyper-growth narrative of the past two years. While “durable” is positive for a long-term investor, it signals “slower” to a market priced for perfection. This deceleration, combined with the margin compression, was the one-two punch that caused the sell-off.
3. The AI Story: Still Bullish, but Not a Margin Saver (Yet)
Management remains extremely bullish on AI but also implicitly confirmed it’s not a high-margin business.
Bullish Targets: The company is on track for its stated goal of $1.5 billion in aggregate AI revenue for 2025.
Open Ecosystem: Management emphasized its open approach, highlighting partnerships and interoperability with a broad range of players beyond just Nvidia, including AMD, Broadcom, and others.
The Margin Dilution: The Q3 commentary, combined with the Q4 guide, confirms that these large-scale AI build-outs are lower-margin than Arista’s traditional enterprise business. Arista is winning massive AI deals, but it comes at the cost of its blended gross margin.
4. Key Strategic Theme: The “Centers of Data” Strategy
You will hear this phrase repeatedly. CEO Jayshree Ullal has reframed Arista’s entire portfolio under this new umbrella.
What it is: This strategy unifies Arista’s separate products (campus, data center, AI) into a single, cohesive architecture. The message is that whether a customer is an enterprise building an office network or a “cloud titan” building an AI cluster, they are all just creating “centers of data.”
Why it Matters: This is a smart marketing and strategic pivot. It allows Arista to:
Cross-sell its entire portfolio.
De-emphasize its reliance on any one segment (like lumpy cloud/AI deals).
Frame its business as a long-term, durable platform rather than a high-beta AI-only play.
Conclusion: Analyst Take
The Q3 call was a “reset” of expectations. Management used the “favorable mix” in Q3 to beat estimates, then used the Q4 guide to bring expectations back to a “normalized” (and lower) baseline. The market, which hates uncertainty, sold off aggressively on the two key negatives: margin compression and growth deceleration.
The bullish case is that this is a temporary normalization and that the AI revenue will become too large to ignore. The bearish case is that this is the new normal—that Arista’s growth is slowing and its margins are permanently impaired by the very AI deals it’s winning.
The whisper number theory makes sense when you consider the stock's valuation multiple had expanded to perfection pricing. A 36% YoY revenue jump with strong AI partnerships should have been enough, but the market clearly wanted more than $9.6B Q4 guidance. It's the classic problem of winning on fundamentals but losing on sentiment becuase expectations got too detached from reality.