I Sold My Warner Bros. Shares and Bought These 2 Growth Stocks Instead. The AI Supercycle Made It Obvious.

The article is bullish on Broadcom and AMD as next-wave AI beneficiaries, citing Broadcom’s networking and custom ASIC opportunities and AMD’s inference/agentic AI exposure. Broadcom’s custom chip sales are projected to reach $100 billion in fiscal 2027, while AMD has secured commitments for 6 gigawatts of GPUs each from OpenAI and Meta. The piece is investment commentary rather than new company disclosure, but it highlights a constructive long-term growth thesis for both stocks.

Analysis

The market is likely underpricing how AI infrastructure spend is shifting from a pure compute race to a systems bottleneck race. That favors AVGO and AMD because their economics improve as cluster size rises: networking, custom silicon, memory density, and rack-level integration all become higher-value content per deployed dollar. The second-order loser is commodity server and legacy networking spend — if hyperscalers increasingly design around integrated rack solutions and custom ASICs, margin pools migrate away from generic accelerators and off-the-shelf hardware.

AVGO has the cleaner near-term monetization path because custom silicon and Ethernet switching are already embedded in hyperscaler capex, while AMD’s payoff is more execution-dependent but potentially more asymmetric. The real option value in AMD is not just the GPU product cycle; it is the attach rate from CPUs, software, and full-rack systems if inference clusters become more CPU-heavy than today’s training-dominated configurations. That creates a multi-year share gain opportunity, but also raises the bar for supply-chain execution and software adoption.

A contrarian read: consensus is focusing on the size of the AI capex pie, but the key question is allocation efficiency. If model economics improve, hyperscalers may delay some of the most aggressive hardware spending while extracting more output from existing clusters, which would be a headwind to the most crowded AI beneficiaries over the next 1-2 quarters. Conversely, if agentic AI drives persistent inference demand, the demand curve could steepen again in 2H, with AVGO likely the earliest beneficiary and AMD the more levered catch-up trade.

Key Decisions for Investors

• Long AVGO on a 6-12 month horizon; highest-quality exposure to AI networking/custom silicon with the clearest visibility into hyperscaler design wins. Favor pullbacks versus chasing strength; risk/reward is attractive if the market continues to re-rate AI infrastructure from chip unit growth to content-per-rack growth.
• Long AMD vs. short NVDA in a 3-6 month pair trade if you want to express the inference/CPU mix shift rather than broad AI beta. The thesis is that AMD’s upside is more underappreciated, while NVDA is more exposed to any pause in incremental training-heavy capex.
• Buy AVGO / short INTC as a structural pair over 12 months. AVGO benefits from the move toward custom silicon and networking intensity; INTC remains a laggard in the next-gen AI rack architecture unless it proves meaningful design-win conversion.
• For tactical exposure, use AMD Jan-2027 call spreads instead of common stock if you want convexity on execution. The market is likely to punish any delay in ROCm adoption or rack-scale rollout, so defined risk is preferable to outright size.
• Avoid overcommitting to the theme via META and GOOGL here; they are beneficiaries but not pure plays. If hyperscaler capex discipline tightens, their stocks can de-rate even while AVGO and AMD still gain share.