Which Semiconductor Stock Looks Best Right Now?
AI is the future. But here’s the truth most headlines ignore: AI doesn’t exist without chips.
Behind every ChatGPT query, image generation, autonomous vehicle, or AI-powered search engine are semiconductors—specifically, high-performance GPUs and accelerators that can process massive volumes of data in real time.
That’s why, as an investor, I believe semiconductors are the most important infrastructure play of this entire technological wave. The opportunity is massive—and it’s already underway.
In this post, I’ll break down the three key players in this space—Nvidia, AMD, and Intel—and share:
A financial snapshot of each company
Their AI positioning and competitive edge
Risks and macro headwinds
My portfolio allocation view for this sector
And a legally compliant conclusion
🎁 START YOUR 30-DAY FREE TRIAL NOW
Important: On iPhone, you must tap this link — Substack’s subscribe button won’t show the free trial.
👉 Start your free trial
Nvidia (NVDA): The Undisputed Leader
Nvidia is the engine of AI. Over 80% of AI training workloads globally are run on their GPUs. From OpenAI to Microsoft Azure to Tesla’s Dojo system—if it’s AI, there’s a good chance it runs on Nvidia.
Financials
Q1 FY2026 revenue: $44.1B (up 69% YoY)
Data Center revenue: $39.1B (up 73%) — now ~90% of total revenue
Net income: $14.88B this quarter, over $72.9B for FY2025
Gross margin: 70%
Cash & equivalents: $31.4B
P/E ratio (forward): ~50x
Nvidia’s profitability is outstanding. Few companies combine growth and margin strength like this.
AI Advantage
Their Blackwell architecture chips are sold out through 2025.
CUDA, cuDNN, and software stack keep developers locked into their platform.
Ecosystem partnerships with AWS, Google Cloud, Microsoft, and Meta make it hard to switch.
What Sets Them Apart
AI stack dominance: From hardware to software.
Scale moat: Economies of scale, capacity pre-booked.
Speed of innovation: Multiple new architectures in the pipeline (e.g. Rubin in 2026).
AMD (AMD): The Fast-Moving Challenger
AMD is the underdog—but they’re closing the gap fast. Their strategy? Deliver performance per dollar. Their MI300X and upcoming MI355 chips compete directly with Nvidia’s H100 and B200 chips.
Financials
Q1 2025 revenue: $7.438B (up 36% YoY)
Data Center revenue: $3.7B (up 57%)
Gross margin: 54%
Operating margin: 24%
P/E ratio (forward): ~24x
While AMD is smaller than Nvidia, it's growing faster in data center. Their margins are improving as they scale production.
AI Advantage
AMD’s MI300 series reportedly offers ~40% more tokens per dollar than Nvidia’s B200—huge for cost-conscious buyers.
Full-stack rack solutions (via ZT Systems acquisition) increase TAM.
Clients include Meta, Microsoft, Oracle, OpenAI.
What Sets Them Apart
More affordable AI compute for scale-out deployments.
Full-rack integration: Not just chips, but turnkey AI servers.
Strategic acquisitions to build data center muscle fast.
Intel (INTC): Legacy Giant in Transition
Intel is trying to re-enter the race. While dominant in CPUs, it has fallen behind in AI hardware. Still, they’re investing heavily to catch up—with a multi-year roadmap that includes Gaudi AI accelerators and their foundry business.
Financials
Current share price: ~$21.88
FY2024 revenue: ~$54.2B (flat)
Gross margin: ~40% (lower than peers)
Cash: $7.8B
P/E ratio (forward): ~10x
Intel is cheap—because the market has little faith in its near-term AI competitiveness. However, it still generates strong cash flow and owns its own fabs (a rare advantage).
AI Advantage (Still Unproven)
Their Gaudi accelerators are 2x–4x cheaper per token in some tests, but adoption is weak.
Intel is now providing foundry services for other chipmakers (including AI startups).
Partnered with Hugging Face and Stability AI—but far behind in traction.
What Might Save Them
U.S. government subsidies and defense contracts
Huge capacity and geographic diversity
Rebuilding credibility under CEO Pat Gelsinger
Key Risk Scenarios to Watch
Even with strong tailwinds, this sector has real risks. Here are the biggest ones:
1. Geopolitical Restrictions
Nvidia and AMD both face U.S. export controls limiting shipments to China. Nvidia’s last quarter included a $4.5B inventory write-down tied to unsold chips for that market. Future restrictions could tighten.
2. AI Hype Cycle Cooldown
If AI spending slows or flattens due to macroeconomic conditions, chip orders could fall short. Nvidia’s valuation leaves little margin for error here.
3. New Entrants or Open Standards
Startups (like Cerebras, Tenstorrent, Groq) or open-source AI chip designs could undercut existing leaders. Nvidia’s moat is strong but not untouchable.
4. Manufacturing Bottlenecks
Most advanced chips are made by TSMC in Taiwan. Any political instability or capacity issues there could disrupt the supply chain.
Portfolio Strategy & Recommendations
Here’s how I’d personally think about allocating capital across these three companies, based on risk, return potential, and AI exposure.
CompanyTypeAllocationRiskUpsideNvidiaMarket leader50–60%MediumModerate but steadyAMDChallenger30–40%Medium–HighHighIntelSpeculative0–10%HighLow–Moderate
My Reasoning:
Nvidia is the blue-chip in this space. Dominant margins, proven products, predictable roadmap.
AMD is the high-beta play. It could outperform percentage-wise as more companies look for affordable AI scaling.
Intel is cheap but risky. A turnaround could pay off, but there’s no strong signal yet.
This spread gives exposure to both stability and upside while managing downside risk.
Looking Ahead: What the Future Could Hold
The semiconductor market is entering a new era:
AI demand is just beginning. Training large models will remain compute-heavy, and inference (running models) at scale will demand even more chips.
Smaller companies and countries will adopt AI. Meaning the TAM is global and growing.
Custom chips and open models will emerge, but they’ll still require GPUs to run efficiently.
More vertical integration: Cloud providers may build their own chips (like Amazon’s Trainium), but Nvidia’s ecosystem advantage will remain for years.
If AI becomes what many expect it to be — a foundational layer of all industries — then the companies building the infrastructure will be among the biggest winners of the decade.
I see this sector as one of the strongest long-term investment opportunities of the next 5–10 years. As with all new technologies, there will be winners, losers, and surprises along the way. But in terms of infrastructure and necessity, AI chips are non-negotiable.
My personal view:
Nvidia remains a core holding — dominant, proven, but not cheap.
AMD is the momentum play — lean, aggressive, and closing in.
Intel is the wild card — potentially a value buy if they can show real progress.
⚖️ Disclaimer (Please Read)
This post reflects my personal opinions and research. It is for informational and educational purposes only. It is not financial advice. I am not a licensed investment advisor, and nothing in this newsletter should be considered a recommendation to buy, sell, or hold any security. Investing involves risk, including the loss of capital. Always consult with a certified financial professional and do your own due diligence before making any investment decisions.