One stocks I am buying now ( watch before 2 late)

Mar 11, 2025

3 min read

Updated: Nov 13, 2025

Advanced Micro Devices

While Advanced Micro Devices (NASDAQ: AMD) remains the No. 2 player in GPUs behind Nvidia—with roughly 10% market share—the company has quietly executed exceptionally well in the CPU market, especially inside data centers. Nvidia may dominate the “power” side of AI with GPUs, but CPUs remain the brains of the operation. And AMD’s progress here has been nothing short of impressive.

The CPU market may not be as large as the GPU market, but it is still expanding rapidly as AI infrastructure buildouts continue. Last quarter, AMD reported that its EPYC CPUs now hold well over 50% market share among hyperscalers, who run the largest data centers in the world. EPYC is also gaining consumer momentum, capturing strong share across Amazon, Newegg, MindFactory, and other retail PC platforms.

On the GPU side, AMD is still very much in the game. Microsoft and Meta are already using its MI300X accelerators, and demand for the upcoming MI350 series is strong. The company plans to launch the MI400 family in 2026. Yes, Nvidia still has a major lead in software—CUDA is a giant moat—and that gap will likely keep AMD in second place. But AMD is carving out a growing niche in the inference market, which is scaling rapidly and becoming an increasingly important portion of AI workloads.

The “forced to buy Nvidia” era won’t last forever

Over the last year, investors have acted like Nvidia is the only legitimate choice. In 2024, people were calling AMD “the next Nvidia,” and the stock shot to $230. But in 2025, sentiment flipped dramatically—now people compare AMD to Intel, which is honestly ridiculous given AMD’s performance and roadmap.

The reality is simple:Today, companies feel pressured to buy Nvidia because of the CUDA ecosystem and because everyone wants to avoid falling behind. But I don’t believe this dynamic will hold over the next 3–5 years. As AI becomes more standardized and cost-sensitive, the market will naturally migrate toward cheaper, flexible, high-performance alternatives—and this is exactly where AMD shines.

With low-cost AI chips emerging in China, and global enterprises facing rising AI infrastructure bills, AMD’s more affordable GPU offerings become extremely attractive. This will likely open the door for companies to diversify hardware providers rather than relying on Nvidia alone. The AI chip market will not be winner-takes-all—it will be multi-vendor, cost-driven, and performance balanced, benefiting AMD significantly.

Valuation: Forward P/E ~15 vs. S&P 500 at ~22 — absurdly cheap

At a forward P/E near 15, AMD is shockingly inexpensive for the second-largest AI chip company in the world. For perspective, the S&P 500 trades at around 22x forward earnings, meaning AMD is cheaper than the average U.S. company, despite its huge positioning in AI, data centers, and CPUs.

Dig deeper, and the value becomes even more obvious:

50%+ gross margins
Expanding data center revenue
Growing GPU adoption
Strong product roadmap (MI350 → MI400)
EPYC dominance among hyperscalers

This valuation doesn’t match the company’s strategic position at all. In fact, it’s almost diabolically cheap.

And based on my long-term cash flow projections, the future P/E (5–10 year horizon) at today’s price (~$82 at the time of writing) could realistically land in the 10–20 range. For a company of AMD’s quality, growth rate, and market relevance, that is unbelievably undervalued.

Long-Term Thesis: Data center CPUs will be AMD’s true revenue machine

Most investors focus entirely on GPUs. But I believe AMD’s biggest long-term driver will be data center CPUs.

EPYC has already taken massive market share.
AI infrastructure still requires powerful CPUs.
Hyperscalers prefer mixing vendors for cost optimization.
AMD is attacking both inference GPUs and data center CPUs simultaneously.

GPUs will absolutely contribute to growth—but over the next decade, data center CPUs will likely be the foundation of AMD’s revenue engine, with GPUs adding upside optionality.

With AI spending climbing and hyperscalers expanding at a rapid pace, AMD’s runway is extremely long.

Conclusion

AMD today trades like a forgotten AI company, priced at a discount more suited for a slow-growth legacy provider. But the fundamentals tell a different story: expanding market share, a strong product roadmap, CPU dominance in hyperscalers, and increasing GPU adoption across big tech.

With a forward P/E of ~15, expanding margins, and multiple growth vectors, AMD’s valuation looks increasingly attractive. If sentiment shifts even slightly—or if GPU adoption accelerates—the stock has enormous upside.

KEEP BUYING AMD and DON'T STOP BUYING