which semiconductor stocks should i consider?

Top semiconductor stocks to consider include market leaders driven by artificial intelligence hardware demand, custom silicon, and essential manufacturing infrastructure like Nvidia (NVDA), Taiwan Semiconductor Manufacturing Co. (TSM), and Broadcom (AVGO).

Top Semiconductor Stocks to Watch

• Nvidia Corp. (NVDA) : Dominates the market for artificial intelligence graphics processing units (GPUs) and accelerated computing infrastructure.
• Taiwan Semiconductor Manufacturing Co. Ltd. (TSM) : Operates as the world's leading pure-play semiconductor foundry, manufacturing advanced chips for major designers.
• Broadcom Inc. (AVGO) : Excels in custom AI accelerators (ASICs) and high-speed networking switches essential for data center clusters.
• Micron Technology Inc. (MU) : Benefits from massive demand for high-bandwidth memory (HBM) required by modern AI workloads.
• Advanced Micro Devices Inc. (AMD) : Competes strongly in high-performance processors, GPUs, and enterprise data center solutions.
• Applied Materials Inc. (AMAT) : Provides critical manufacturing equipment, services, and software for the fabrication of advanced microchips.
• ASML Holding NV (ASML) : Holds a near-monopoly on extreme ultraviolet (EUV) lithography systems required to print the smallest, most advanced semiconductors.

If you'd like to narrow this down, let me know:

• Your primary investment goal (growth, dividends, or risk management)
• Your preferred holding period
• Whether you prefer individual stocks or broad semiconductor ETFs

I can tailor a specific approach to fit your portfolio strategy.

If you're looking at semiconductor stocks in 2026, I'd focus on companies with durable exposure to AI infrastructure rather than simply buying the fastest-growing chip stock.

The backdrop is unusually strong: Omdia expects semiconductor revenue to rise sharply in 2026, with AI-driven HBM/memory demand and advanced packaging remaining supply-constrained into at least 2027.

My semiconductor shortlist

1. NVIDIA — best overall AI semiconductor play

NVIDIA is still my first name to investigate.

Its latest results were extremely strong: $96B quarterly revenue and guidance for $108B the following quarter, while management's longer-term outlook implies exceptionally high growth.

The important thing isn't just GPUs. NVIDIA has built an ecosystem around networking, systems, CUDA/software and increasingly rack-scale infrastructure. Gartner currently identifies NVIDIA as the leader in AI network fabric.

Main concern: expectations are enormous. A great company can still be a mediocre investment if you overpay.

2. TSM — my favorite "picks and shovels" choice

Taiwan Semiconductor Manufacturing Company gives you exposure to the semiconductor industry without having to predict which chip designer wins.

TSMC manufactures chips for NVIDIA, AMD, Apple and many other leading companies. Advanced packaging is also becoming a major AI bottleneck, with TSMC's advanced-packaging capacity reportedly heavily utilized.

Why I like it: you're essentially betting that demand for leading-edge compute continues, regardless of whether NVIDIA, AMD, custom ASICs or another architecture captures the revenue.

Biggest risk: Taiwan/China geopolitics.

3. AVGO — potentially the most interesting alternative to NVIDIA

Broadcom is particularly attractive because of custom AI silicon.

Hyperscalers increasingly want their own AI accelerators, and Gartner identifies Broadcom as the current leader in custom AI silicon.

That gives Broadcom a different AI opportunity from NVIDIA: rather than betting entirely on one dominant GPU platform, you're betting on the growth of customized AI infrastructure.

Broadcom reports next week, so there's also a near-term catalyst—but that means potentially significant earnings volatility.

4. ASML — the semiconductor industry's tollbooth

ASML is one of my favorite long-term semiconductor businesses.

It makes the extremely sophisticated lithography equipment necessary for producing the world's most advanced chips. In other words, NVIDIA and AMD can compete with each other, but both need increasingly advanced manufacturing equipment.

ASML also benefited from the latest NVIDIA-driven semiconductor rally.

The downside is valuation and geopolitical/export-control risk.

5. Micron — higher-risk, potentially higher-reward

Micron Technology is the one I'd consider if you want more aggressive exposure.

AI accelerators require huge quantities of HBM (high-bandwidth memory). Omdia expects memory to represent more than half of semiconductor revenue in 2026, with HBM supply remaining constrained.

The catch: memory is notoriously cyclical. Micron has also already experienced an enormous run, so I'd be much more valuation-sensitive here than with TSMC.

6. AMD — compelling, but I'd rank it below NVDA

Advanced Micro Devices has a very interesting position in data-center CPUs and AI accelerators. Gartner currently identifies AMD as the leader in enterprise AI server CPUs.

The problem is that AMD stock has already risen roughly 120%+ in 2026, meaning a lot of optimism is already embedded in the price.

I'd happily own AMD, but I'd want a better entry price than I'd require for TSMC.

If I were building a semiconductor basket

For a long-term, 5–10 year portfolio, I'd roughly think about it this way:

• 25% NVDA — AI compute leader
• 20% TSM — manufacturing/foundry
• 15% AVGO — custom AI + networking
• 15% ASML — semiconductor equipment
• 10% MU — HBM/memory
• 10% AMD — AI/data-center challenger
• 5% LRCX or AMAT — semiconductor equipment

I wouldn't necessarily buy all of them today. Entry price matters enormously, especially after the AI-driven semiconductor rally.

One interesting current signal: NVIDIA's latest forecast helped revive the entire chip sector, but some names are behaving very differently. For example, Marvell fell sharply after its latest results despite beating estimates, illustrating how high expectations have become for AI-related semiconductor stocks.

If you tell me your investment horizon (e.g. 2, 5, or 10 years) and whether you're comfortable with high volatility, I can narrow this to the 3 semiconductor stocks I'd consider buying at today's prices and compare their valuations/upside.

Its latest results were extremely strong: $96B quarterly revenue and guidance for $108B the following quarter, while management's longer-term outlook implies exceptionally high growth.

The important thing isn't just GPUs. NVIDIA has built an ecosystem around networking, systems, CUDA/software and increasingly rack-scale infrastructure. Gartner currently identifies NVIDIA as the leader in AI network fabric.

Main concern: expectations are enormous. A great company can still be a mediocre investment if you overpay.

2. TSM — my favorite "picks and shovels" choice

Taiwan Semiconductor Manufacturing Company gives you exposure to the semiconductor industry without having to predict which chip designer wins.

TSMC manufactures chips for NVIDIA, AMD, Apple and many other leading companies. Advanced packaging is also becoming a major AI bottleneck, with TSMC's advanced-packaging capacity reportedly heavily utilized.

Why I like it: you're essentially betting that demand for leading-edge compute continues, regardless of whether NVIDIA, AMD, custom ASICs or another architecture captures the revenue.

Biggest risk: Taiwan/China geopolitics.

3. AVGO — potentially the most interesting alternative to NVIDIA

Broadcom is particularly attractive because of custom AI silicon.

Hyperscalers increasingly want their own AI accelerators, and Gartner identifies Broadcom as the current leader in custom AI silicon.

That gives Broadcom a different AI opportunity from NVIDIA: rather than betting entirely on one dominant GPU platform, you're betting on the growth of customized AI infrastructure.

Broadcom reports next week, so there's also a near-term catalyst—but that means potentially significant earnings volatility.

4. ASML — the semiconductor industry's tollbooth

ASML is one of my favorite long-term semiconductor businesses.

It makes the extremely sophisticated lithography equipment necessary for producing the world's most advanced chips. In other words, NVIDIA and AMD can compete with each other, but both need increasingly advanced manufacturing equipment.

ASML also benefited from the latest NVIDIA-driven semiconductor rally.

The downside is valuation and geopolitical/export-control risk.

5. Micron — higher-risk, potentially higher-reward

Micron Technology is the one I'd consider if you want more aggressive exposure.

AI accelerators require huge quantities of HBM (high-bandwidth memory). Omdia expects memory to represent more than half of semiconductor revenue in 2026, with HBM supply remaining constrained.

The catch: memory is notoriously cyclical. Micron has also already experienced an enormous run, so I'd be much more valuation-sensitive here than with TSMC.

6. AMD — compelling, but I'd rank it below NVDA

Advanced Micro Devices has a very interesting position in data-center CPUs and AI accelerators. Gartner currently identifies AMD as the leader in enterprise AI server CPUs.

The problem is that AMD stock has already risen roughly 120%+ in 2026, meaning a lot of optimism is already embedded in the price.

If you're looking for semiconductor stocks to research in 2026, I'd focus less on trying to pick the single "best" chip company and more on owning different parts of the semiconductor value chain. The industry backdrop is unusually strong: Gartner projects semiconductor revenue above $1.3 trillion in 2026, while Omdia expects AI-driven demand to keep HBM, advanced packaging, and leading-edge capacity constrained into 2027.

My shortlist

How I'd rank them

1. TSM — my favorite "picks and shovels" semiconductor play. TSMC manufactures chips for NVIDIA, AMD, Broadcom and many other major designers. Its advanced packaging and leading-edge manufacturing capacity are particularly valuable while AI demand is overwhelming supply. JPMorgan notes that NVIDIA, Broadcom and AMD collectively spend a substantial portion of their costs with TSMC.

2. NVDA — highest-quality AI growth story, but expectations are enormous. NVIDIA remains the central supplier of AI accelerators, with its next-generation Vera Rubin platform approaching commercialization. The catch is valuation/expectations: the market is demanding extraordinary growth, and NVIDIA reports earnings tomorrow, August 26, making near-term volatility particularly high. Reuters reports analysts expect roughly $92 billion in quarterly revenue.

3. ASML — perhaps the best long-duration semiconductor infrastructure play. ASML makes the advanced lithography equipment necessary to manufacture the world's most sophisticated chips. Its recent results showed strong growth, and it expects capacity expansion in 2027–28.

4. AVGO — excellent way to diversify away from NVIDIA GPUs. Broadcom is positioned around custom AI accelerators and networking. Recent market commentary points to strong hyperscaler spending on custom AI chips.

5. AMD — higher-risk alternative to NVIDIA. AMD gives you exposure to both AI accelerators and server CPUs. It's already had a huge 2026 run, though, so I'd be particularly sensitive to valuation rather than simply buying because the business is growing. A Raymond James analyst upgraded AMD to Strong Buy today, citing the potential for rapid server-CPU growth from AI workloads.

6. MU — the interesting "second-order" AI bet. AI isn't just about GPUs. High-bandwidth memory (HBM) is becoming a major bottleneck, and Micron is one of only three companies producing HBM at scale alongside Samsung and SK Hynix.

7. MRVL — speculative but interesting. Marvell gives you exposure to custom silicon and data-center networking. It's potentially attractive if hyperscalers continue moving toward customized AI infrastructure, but I'd treat it as a smaller position because expectations and volatility are high.

If it were my money

For a 5–10 year semiconductor portfolio, I'd rather have something like:

• 25% TSM
• 20% NVDA
• 15% ASML
• 15% AVGO
• 10% AMD
• 10% MU
• 5% MRVL

That's deliberately diversified across AI accelerators → foundry → lithography → networking/custom silicon → CPUs → memory.

One important caveat: semiconductor stocks have already had a tremendous run, and expectations are extremely high. Recent market commentary specifically notes that strong earnings haven't always translated into stock gains because valuations have become demanding.

If you tell me (1) how much you want to invest, (2) your time horizon, and (3) whether you're conservative, moderate, or aggressive, I can narrow this to 3–5 stocks and suggest reasonable entry/position sizes rather than just giving you a semiconductor watchlist.

3. ASML — perhaps the best long-duration semiconductor infrastructure play. ASML makes the advanced lithography equipment necessary to manufacture the world's most sophisticated chips. Its recent results showed strong growth, and it expects capacity expansion in 2027–28.

4. AVGO — excellent way to diversify away from NVIDIA GPUs. Broadcom is positioned around custom AI accelerators and networking. Recent market commentary points to strong hyperscaler spending on custom AI chips.

5. AMD — higher-risk alternative to NVIDIA. AMD gives you exposure to both AI accelerators and server CPUs. It's already had a huge 2026 run, though, so I'd be particularly sensitive to valuation rather than simply buying because the business is growing. A Raymond James analyst upgraded AMD to Strong Buy today, citing the potential for rapid server-CPU growth from AI workloads.

6. MU — the interesting "second-order" AI bet. AI isn't just about GPUs. High-bandwidth memory (HBM) is becoming a major bottleneck, and Micron is one of only three companies producing HBM at scale alongside Samsung and SK Hynix.

7. MRVL — speculative but interesting. Marvell gives you exposure to custom silicon and data-center networking. It's potentially attractive if hyperscalers continue moving toward customized AI infrastructure, but I'd treat it as a smaller position because expectations and volatility are high.

If it were my money

For a 5–10 year semiconductor portfolio, I'd rather have something like:

• 25% TSM
• 20% NVDA
• 15% ASML
• 15% AVGO
• 10% AMD
• 10% MU
• 5% MRVL

That's deliberately diversified across AI accelerators → foundry → lithography → networking/custom silicon → CPUs → memory.

One important caveat: semiconductor stocks have already had a tremendous run, and expectations are extremely high. Recent market commentary specifically notes that strong earnings haven't always translated into stock gains because valuations have become demanding.

The semiconductor sector continues to experience massive momentum driven by relentless AI infrastructure demand, data center expansion, and complex hardware architectures.

Leading market analysts and financial platforms highlight several key semiconductor stocks across different segments of the chip ecosystem:

• Nvidia Corp. (NVDA) : Remains the dominant powerhouse in AI GPUs and accelerated computing, capturing a massive share of the hardware demand driving modern data centers.
• Broadcom Inc. (AVGO) : A critical "pick-and-shovel" play excelling in custom AI accelerators (ASICs) for hyperscalers like Alphabet and Meta, alongside high-speed networking and switching infrastructure.
• Taiwan Semiconductor Manufacturing Co. (TSM) : The premier global foundry manufacturing advanced nodes for nearly all major fabless chip designers, making it a foundational play on overall industry growth.
• Advanced Micro Devices Inc. (AMD) : Continues to carve out a strong alternative position in high-performance computing, AI accelerators, and data center ecosystems.
• Micron Technology Inc. (MU) : A primary beneficiary of the surging demand for high-bandwidth memory (HBM) required by modern AI processors.
• Applied Materials Inc. (AMAT) & KLA Corp. (KLAC) : Essential semiconductor equipment and testing providers benefiting from the increasing complexity and precision required to manufacture advanced chips.

Note: This overview is for informational purposes and does not constitute formal financial advice. Market conditions change rapidly.

To help narrow down these choices, tell me:

• Are you looking for growth (aggressive AI plays) or value/dividends?
• What is your risk tolerance and investment timeframe?

When considering semiconductor stocks, the market is heavily shaped by the massive infrastructure buildout for artificial intelligence, data centers, and advanced manufacturing . Rather than looking at the sector as a single block, it helps to break down the top considerations by their specific role in the supply chain.

• Nvidia Corporation (NVDA) — The dominant engine of the AI hardware trade. Nvidia continues to command a massive market share in AI accelerators and high-performance GPUs, paired with staggering data center revenue growth.
• Taiwan Semiconductor Manufacturing Co. (TSM) — The premier "foundry" that physically manufactures cutting-edge logic chips for designers like Nvidia and Apple. Their dominant grip on advanced nodes makes them an indispensable backbone of global chip production.
• Broadcom Inc. (AVGO) — A powerhouse in custom AI accelerators (ASICs) for major tech hyperscalers looking to build proprietary chips, alongside market dominance in high-speed data center networking and connectivity silicon.
• Micron Technology (MU) — A standout leader in memory and storage, particularly High Bandwidth Memory (HBM) which is critically required alongside advanced GPUs to prevent performance bottlenecks in AI servers.
• Advanced Micro Devices (AMD) — Nvidia’s primary large-scale challenger in high-performance GPUs, data center CPUs, and expanding AI accelerators.
• ASML Holding (ASML) — The essential "pick-and-shovel" monopoly supplier of extreme ultraviolet (EUV) lithography systems required by foundries to print the world's smallest, most advanced chips.
• Applied Materials (AMAT) & Lam Research (LRCX) — Essential wafer fabrication equipment providers that enable the physical creation, etching, and layering of complex semiconductor architectures.

If you prefer broad exposure to the entire sector rather than picking individual winners, you can also look into diversified exchange-traded funds like the VanEck Semiconductor ETF (SMH).

To help narrow down which options fit your portfolio, let me know:

• Are you looking for high-growth AI plays, stable industry infrastructure , or diversified ETFs?
• What is your view on near-term market volatility versus long-term holding?

If you're looking at semiconductors for a multi-year investment, I'd focus less on finding the single "best" chip stock and more on owning different parts of the AI semiconductor ecosystem. The industry backdrop remains strong: 93% of semiconductor executives surveyed by KPMG expect revenue growth in 2026, while memory/HBM has emerged as one of the industry's biggest growth opportunities.

My semiconductor shortlist

My top five to research first:

1. Nvidia (NVDA) — Still the highest-quality pure AI semiconductor exposure. The major risk is that expectations and valuation are extremely high. Nvidia is scheduled to report its fiscal Q2 results on August 26, so I'd be particularly cautious about buying immediately before earnings.
2. TSMC (TSM) — Probably my favorite "picks-and-shovels" semiconductor stock. It manufactures chips designed by Nvidia, AMD, Apple and others, so you aren't betting on which chip designer ultimately wins. AI demand has been driving record revenue, while capacity constraints have even allowed competitors such as Samsung to raise advanced-foundry prices.
3. Broadcom (AVGO) — An especially interesting alternative to Nvidia. Its AI semiconductor revenue reached $10.8B in Q2 2026, up 143% year over year, and management guided to $16B in Q3. Its exposure to custom AI accelerators and networking gives it a different route to benefiting from hyperscaler AI spending.
4. ASML (ASML) — The "infrastructure behind the infrastructure." Its EUV lithography machines are critical to producing the most advanced chips. This is one of my favorite ways to get semiconductor exposure without betting entirely on GPU competition.
5. Micron (MU) — More cyclical and therefore riskier, but potentially very attractive. AI systems require enormous quantities of high-bandwidth memory, and industry memory investment is expected to exceed $50B in 2026.

How I'd think about them

If you want maximum AI upside: NVDA + AVGO + AMD + MU

If you want higher-quality, somewhat more diversified semiconductor exposure: TSM + ASML + NVDA + AVGO

If you want a more aggressive portfolio: NVDA + AMD + MU + MRVL

If you want less individual-stock risk, I'd also consider a semiconductor ETF such as SOXX rather than trying to pick winners.

One important caveat: semiconductor stocks have been extremely volatile this year. The Philadelphia Semiconductor Index recently experienced a roughly 29% drawdown followed by a rapid rebound, and higher Treasury yields have also caused sharp sector-wide selloffs. Investor's Business Daily Financial Times So I'd be more inclined to build a position gradually than put the entire amount in at once.

If you tell me how much you're looking to invest (e.g. $10k, $50k, $100k) and your time horizon (3, 5, or 10+ years), I can rank these by expected risk/reward and give you a sample semiconductor portfolio.

When considering semiconductor stocks, the market is heavily shaped by the massive infrastructure buildout for artificial intelligence, high-performance computing, and critical hardware bottlenecks.

The top semiconductor players are categorized by their specific role in the supply chain:

• GPU & AI Compute Leaders (The Innovators)
  • Nvidia (NVDA): Remains the dominant driving force of the AI hardware trade with strong data center demand and its latest platform rollouts.
  • Advanced Micro Devices (AMD): Continues closing the gap as a primary alternative for high-performance AI accelerators and enterprise computing.
• Custom Silicon & Networking (The Infrastructure Builders)
  • Broadcom (AVGO): A powerhouse in co-designing custom AI application-specific integrated circuits (ASICs) for major cloud providers alongside dominant networking and switching gear.
• Foundry & Manufacturing (The Foundation)
  • Taiwan Semiconductor Manufacturing Company (TSM): Operates as the near-monopoly bottleneck for physically manufacturing cutting-edge logic and AI chips designed by companies like Nvidia and Apple.
• Memory & Storage (The High-Demand Bottleneck)
  • Micron Technology (MU): Emerging as a massive outperformer due to surging demand for high-bandwidth memory (HBM) stacked right alongside advanced processors.
• Chipmaking Equipment (The Pick-and-Shovel Plays)
  • ASML Holding (ASML): Commands a functional monopoly on extreme ultraviolet (EUV) lithography systems required to print the world's most advanced chips.
  • Applied Materials (AMAT) & Lam Research (LRCX): Essential equipment providers benefiting heavily from advanced packaging and foundry expansions.
• Diversified Exposure (Lower Volatility)
  • If picking individual winners feels too concentrated, you can look at broad sector funds like the VanEck Semiconductor ETF (SMH) or SPDR S&P Semiconductor ETF (XSD).

To help narrow down which options fit your strategy, tell me:

• Are you looking for growth/momentum or value/dividends?
• Do you prefer individual stocks or a diversified ETF approach?

If you're looking at semiconductors for a multi-year investment, I’d focus on companies with durable competitive advantages rather than simply chasing whichever chip stock is hottest.

The semiconductor backdrop remains unusually strong: industry executives expect 2026 revenue growth, with AI and data centers the biggest drivers, while memory/HBM has become one of the industry's strongest growth areas.

My semiconductor shortlist

If I had to narrow it to three: NVDA, TSM, and AVGO.

Why TSM deserves particular attention: its July 2026 revenue jumped 45% year over year, and it raised its 2026 growth outlook to more than 40% in U.S.-dollar terms, driven heavily by AI demand.

For a more aggressive portfolio, I'd add AMD and MU. For a somewhat more diversified way to benefit from the semiconductor supply chain, ASML is particularly interesting because it sells the manufacturing equipment rather than betting on one chip designer. Recent analysis also highlights the complementary roles of Nvidia, TSMC and ASML in the AI semiconductor ecosystem.

One caution: great semiconductor companies aren't automatically great stocks at any price. The sector has rallied sharply from its July lows, and analysts are increasingly warning that expectations and valuations are high even though fundamentals remain strong.

If you tell me (1) how much you want to invest, (2) your time horizon, and (3) whether you're conservative/moderate/aggressive, I can narrow this to a 3–5 stock semiconductor portfolio with suggested percentages and the bull/bear case for each.

When evaluating semiconductor stocks, the industry is heavily driven by massive artificial intelligence infrastructure spending, cloud data center expansions, and tight memory supply chains.

Rather than looking at the sector as a single block, it helps to break top contenders down by their specific role in the tech supply chain:

• Nvidia Corporation (NASDAQ: NVDA) : The undisputed powerhouse of AI compute and GPUs. With the rollout and production ramp of its next-generation platforms, it remains the central engine of the AI trade, though valuations reflect high expectations.
• Taiwan Semiconductor Manufacturing Company (NYSE: TSM) : The global manufacturing backbone. As the foundry that physically builds cutting-edge chips for designers like Nvidia and Apple, TSM benefits from soaring utilization and expansion into advanced 2-nanometer production nodes.
• Broadcom Inc. (NASDAQ: AVGO) : A dual threat in AI silicon. Broadcom wins by designing custom application-specific integrated circuits (ASICs) for major cloud hyperscalers looking to diversify, alongside dominating essential data center networking infrastructure.
• Micron Technology, Inc. (NASDAQ: MU) : The standout performer in high-bandwidth memory (HBM). Surging requirements for fast memory stacked alongside AI accelerators have created exceptional pricing power and revenue growth for memory leaders.
• Advanced Micro Devices, Inc. (NASDAQ: AMD) : A major competitor in high-performance CPUs and a growing alternative for AI accelerators. As data centers scale out multi-vendor rack architectures, AMD captures significant enterprise server share.
• ASML Holding N.V. (NASDAQ: ASML) : The "pick-and-shovel" play. Holding a crucial technological edge in lithography equipment required to print the world's most advanced chips, it is deeply tied to long-term global fab buildouts.

To help narrow down which choices fit your strategy, tell me:

• Are you looking for aggressive growth, dividends/income , or diversification via ETFs (like VanEck Semiconductor ETF)?
• Do you have a specific risk tolerance regarding valuation volatility?

If you’re looking at semiconductors today, I’d focus on a handful of high-quality names rather than trying to own the whole sector. The AI buildout remains a major tailwind, but valuations and expectations are also elevated, so entry price matters.

My semiconductor shortlist

1. NVIDIA (NVDA) — Still my first choice for pure AI-semiconductor exposure. Its moat in accelerated computing is exceptional, but the stock now carries a very large valuation and therefore requires continued extraordinary growth.

2. TSMC (TSM) — Probably my favorite risk/reward semiconductor holding. TSMC manufactures chips for NVIDIA, AMD, Apple and others, so you aren't betting on a single chip designer winning. Recent results have also supported a strong 2026 outlook. The obvious risk is Taiwan geopolitical exposure.

3. Broadcom (AVGO) — Particularly interesting because AI infrastructure isn't just GPUs. Broadcom participates in custom AI accelerators and networking, giving you exposure to hyperscaler spending from a different angle.

4. ASML (ASML) — A fascinating long-term semiconductor infrastructure play. Its lithography technology is critical to manufacturing leading-edge chips, meaning it benefits from continued fab investment regardless of which AI chip designer wins.

5. Micron (MU) — More cyclical and therefore riskier, but potentially very attractive if AI-driven demand for high-bandwidth memory remains strong. I'd consider it a satellite position rather than the core holding.

6. AMD (AMD) — The higher-risk alternative to NVIDIA. If AMD takes more AI accelerator share, the upside could be substantial, but you're betting on a much more competitive position.

If it were my money

For a 5–10 year semiconductor basket, I'd roughly think about:

• 30% NVDA
• 25% TSM
• 20% AVGO
• 15% ASML
• 10% MU/AMD

I wouldn't necessarily buy all at once. The semiconductor sector has rallied substantially and analysts are explicitly warning that expectations may be getting ahead of fundamentals.

If you tell me (1) how much you want to invest, (2) your time horizon, and (3) whether you prefer aggressive growth or lower risk, I can narrow this to 3–5 stocks and give you specific buy ranges/valuation levels to watch.