Chip stocks set for secular upside as AI demand surges
Bank of America analyst Vivek Arya identifies seven chip stocks, including Nvidia, poised for gains as AI demand outpaces supply. The AI data center gear market is projected to reach $1.7 trillion by 2030. While hyperscaler spending may leave little spare cash by 2027, BofA believes demand will sustain investment through 2027.

*this image is generated using AI for illustrative purposes only.
Bank of America sees the semiconductor sector as one of the most powerful secular calls for sustained outperformance in the AI era, driven by scarcity rather than financing risks. In a note published Monday, BofA Securities analyst Vivek Arya wrote that the focus has shifted from whether AI spending will generate returns to whether chipmakers can keep up with demand. Despite volatility from geopolitics and interest rates, the firm believes scarcity drives secular upside for chip stocks.
AI demand and market projections
AI usage has surged, with token consumption up roughly sevenfold in a year and Google AI queries rising more than 300%. Agentic software, which performs multi-step tasks autonomously, can consume 10 to 1,000 times more tokens than a single chat, creating a steady demand for computing power. Arya dismissed concerns that record cash raising by hyperscalers signals a warning, noting that chip cycles peak on oversupply, not financing. With power, land, wafers, and memory scarce, capacity cannot be overbuilt fast enough to flood the market.
The bank projects the market for AI data center gear will top $1.7 trillion by 2030, up from about $264 billion in 2025. AI chips alone are expected to account for roughly $1.2 trillion of that total. Combined spending by the top five U.S. cloud companies—Alphabet Inc., Microsoft Corp., Amazon.com Inc., Meta Platforms Inc., and Oracle Corp.—should hit about $771 billion next year, an increase of approximately 68%.
Top stock picks and upside potential
Arya named seven stocks to play this theme, all rated Buy. These companies share a setup of solid fundamentals and room to catch up after lagging the broader chip index this year. Nvidia Corp. has the most upside to BofA's target, with a $350 price objective sitting about 71% above the stock's level at the time of the note.
| Company | Ticker | Upside to Price Target |
|---|---|---|
| Nvidia Corp. | NVDA | 71% |
| Microchip Technology Inc. | MCHP | 38% |
| Texas Instruments Inc. | TXN | 30% |
| Credo Technology Group Holding Ltd | CRDO | 22% |
| Analog Devices Inc. | ADI | 15% |
| KLA Corp. | KLAC | 9% |
| Cadence Design Systems Inc. | CDNS | 6% |
Risks and financing outlook
The boom faces a weak spot as hyperscalers are expected to spend almost every dollar they earn on data centers—95% to 100% through 2028, compared to a third to a half in the past—leaving almost no spare cash by 2027. OpenAI's revenue is projected to climb from $13 billion in 2025 to $283 billion by 2030, though losses may also rise as cumulative computing spend reaches about $665 billion over the period. Arya emphasized that the key risk is not financing, but whether demand continues to absorb incremental supply as capacity expands. BofA expects spending to hold up through 2027, funded by generated cash and long-term contracts.
How will the projected scarcity of power and land impact the geographical distribution of new data center construction?
What are the potential margin implications for hyperscalers if data center spending remains at 95-100% of earnings through 2028?
Could the rapid rise in agentic software usage accelerate the projected $1.7 trillion market size beyond 2030?
























