Big Tech earnings reinforce strength of AI demand across chips, cloud and software: Wedbush

Wedbush says AI demand remains strong across chips, cloud, software and cybersecurity, with Nvidia’s Blackwell and Rubin chip demand/supply imbalance described as "overwhelming bullish." The firm sees accelerating enterprise AI adoption at Microsoft Azure and AWS, cites robust results from AMD, Palantir, Datadog, Snowflake, ServiceNow and Innodata, and forecasts tech stocks could rise another 10% to 12% by year-end. Cybersecurity spend is also expected to increase as AI agents expand demand for endpoint protection, identity governance and cloud security.

Analysis

The biggest second-order effect is not just stronger AI capex, but the widening gap between AI infrastructure winners and the rest of software. When hyperscalers keep raising spend, the near-term beneficiaries are not the application layers that sell the story — it’s the picks-and-shovels stack with pricing power, scarce supply, and multi-quarter backlog visibility. NVDA remains the clearest bottlenecked asset, but the more interesting setup is in AMD and the networking/security adjacency where incremental AI deployments pull through more silicon, observability, identity, and data movement spend.

Consensus is still underappreciating how enterprise deployment changes the mix of demand. The market has treated AI as a one-time model buildout; the next leg is operationalization, which tends to be stickier and broader across departments, so software names tied to workflow automation and telemetry can re-rate even without dramatic revenue inflection today. That argues for a 6-12 month view where the earnings power is still ahead of the visible numbers, particularly for MSFT, AMZN, PLTR, DDOG and NOW.

The contrarian risk is that the market is already paying for perpetual acceleration, so the bar for upside is high if the next print is merely “good.” The more likely failure mode is not demand destruction, but digestion: capex cadence can normalize for a quarter or two, causing high-multiple names to de-rate even if fundamentals remain healthy. On the downside, any evidence of slower AI cloud deal conversion at Azure or AWS would hit the whole complex because it would challenge the idea that enterprise AI is shifting from pilot to production on schedule.

Cybersecurity is the overlooked lever. More AI agents increase the attack surface and the audit burden, which should force budget expansion rather than substitution, especially in identity, endpoint, and cloud workload protection. That makes CRWD, PANW and ZS better expressed as a relative long inside tech than as a pure sector beta trade, because the monetization path is more defensive and less dependent on new model breakthroughs.

Key Decisions for Investors

• Add to NVDA into the next earnings window; use a tight risk box and size for 1-2 quarter convexity, since supply scarcity can extend upside but any guide-down would likely compress the whole AI complex.
• Long AMZN / short a basket of high-multiple, AI-adjacent software names (e.g., SNOW/now with lower conviction names in the group) over 3-6 months; thesis is that cloud monetization outpaces application-layer enthusiasm.
• Buy MSFT on pullbacks for a 6-12 month hold; best risk/reward is as a core AI monetization compounder rather than a trade, with downside cushioned by diversified cash flow.
• Pair long CRWD or PANW vs short a weaker non-AI software name over 1-2 quarters; if AI adoption broadens, security spend should expand faster than general software budgets.
• If you want higher beta exposure, prefer AMD over NVDA only tactically into product-cycle optimism, but keep it paired against a short in a less differentiated semiconductor name to isolate AI server share gains.