If you missed the semiconductor bull run that began in early 2023, you might assume the easy money is gone. You would be wrong. As of the first week of June 2026, the Philadelphia Semiconductor Index (SOX) has seen all of its components rise more than 10% year-to-date, and a wave of AI-driven demand is still accelerating β€” not slowing. Applied Materials CEO Gary Dickerson put it plainly on CNBC on May 28, 2026: "This is the greatest time ever for semiconductors."

But not all semiconductor stocks are equal. Some are priced for perfection with no margin of safety. Some are structurally broken. And a few represent genuine opportunities even at today's elevated prices. This guide covers six major names β€” Nvidia (NVDA), AMD (AMD), Marvell Technology (MRVL), Broadcom (AVGO), Intel (INTC), and Taiwan Semiconductor (TSM) β€” with real prices, analyst targets, and clear buy/avoid/watch ratings for each.

Prices and data are current as of June 2–3, 2026. (Sources: StockAnalysis NVDA, StockAnalysis AMD, StockAnalysis MRVL, StockAnalysis AVGO, StockAnalysis INTC, StockAnalysis TSM)

The Macro Backdrop: Why Semiconductors in 2026

The global semiconductor market has surpassed $700 billion in annual revenue, driven by three converging forces: AI infrastructure buildout, data center expansion, and the early stages of AI-enabled edge devices. Hyperscalers β€” Amazon, Google, Meta, and Microsoft β€” continue to increase capital expenditure at rates that would have seemed unrealistic two years ago. Google alone has committed $80 billion in capex for 2026, a large portion of which flows directly into custom silicon and the chips that power it.

At Computex 2026 in Taipei, Nvidia CEO Jensen Huang called Marvell Technology "the next trillion-dollar company" β€” a statement that sent MRVL shares up 32.52% in a single session on June 2 and added over $65 billion in market cap overnight. That moment crystallized something important: the AI chip ecosystem is not a one-company story. Nvidia leads, but a cohort of companies β€” custom ASIC designers, networking chip makers, and foundries β€” are all scaling rapidly alongside it.

Against this backdrop, here is how the six major names stack up today.

1. Nvidia (NVDA) β€” BUY

Price: $222.82 | 52-week range: $135.40–$236.54 | Market cap: $5.40 trillion | Forward PE: 22.42 | Analyst consensus target: $296.81 (+33%) | Consensus rating: Strong Buy

Nvidia is the closest thing to a must-own stock in this cycle. Its H100 and Blackwell GPU architectures have become the de facto standard for AI training and inference, and the company's CUDA software ecosystem creates switching costs that no competitor has meaningfully cracked. At a forward PE of 22.42 on a company growing revenue at triple-digit rates, the valuation is not expensive β€” it is anomalously cheap relative to growth.

The analyst consensus target of $296.81 implies 33% upside from current levels. That target is not based on optimistic assumptions; it is based on Nvidia's existing product backlog and the contracted spending commitments of its top customers. Every major cloud provider has publicly stated that GPU availability is their primary infrastructure constraint.

The new risk to watch: at Computex, Huang unveiled RTX Spark, an AI PC chip targeting the consumer and enterprise PC segment. This is Nvidia moving downstream into a market currently owned by Intel and AMD. It is bullish for Nvidia's total addressable market and is one more reason the stock's 33% upside estimate may prove conservative.

The buy case in one sentence: Nvidia is the infrastructure layer of the AI economy, priced at a reasonable multiple for its growth rate, with 33% upside to consensus and a dominant competitive moat.

2. Taiwan Semiconductor (TSM) β€” BUY (Best Risk/Reward)

Price: $446.69 | 52-week range: $192.20–$449.39 | Forward PE: 22.89 | Analyst consensus target: $467.84 (+4.7%) | Consensus rating: Strong Buy

TSM is the pick for investors who want semiconductor exposure with the most favorable risk/reward balance. TSMC manufactures chips for virtually every major fabless semiconductor company β€” Nvidia's Blackwell GPUs, Apple's M-series, AMD's EPYC processors, and Marvell's custom AI ASICs all run through TSMC's fabs. There is no viable alternative at leading-edge nodes (3nm, 2nm).

At a forward PE of 22.89, TSM trades at a lower multiple than Nvidia despite being its essential manufacturing partner. The 52-week range tells the story: the stock has more than doubled from its $192.20 low and is now approaching its all-time high near $449. The analyst target of $467.84 suggests modest near-term upside of 4.7%, but that figure is almost certainly stale β€” it does not account for the Nvidia-TSMC AI manufacturing partnership announced this week, which deepens an already critical relationship.

The geopolitical risk (Taiwan Strait tensions) is real but has been priced in and repriced repeatedly. TSMC's Arizona fabs are now operational and receiving US government support. For long-term investors, this is the foundry of the AI age at a reasonable valuation.

The buy case in one sentence: TSM manufactures the chips that run the AI economy, trades at a lower multiple than many of its own customers, and has the lowest downside risk of any name on this list.

3. Broadcom (AVGO) β€” BUY (with Earnings Catalyst)

Price: $481.57 | 52-week range: $241.11–$488.82 | Forward PE: 36.01 | Analyst consensus target: $486.85 | Consensus rating: Strong Buy

Broadcom reports earnings on June 3, 2026 after the market close β€” the day this article is published. The stock is already trading near its all-time high at $481.57, just $7.25 below its 52-week peak of $488.82. The consensus analyst target of $486.85 implies the stock is essentially fairly valued at current prices. So why is it a buy?

Because Broadcom's AI revenue has been growing at a rate that is consistently outrunning analyst models. The company's custom AI ASIC business β€” building bespoke chips for Google's TPU program and other hyperscaler custom silicon initiatives β€” is reported to be generating over $4.1 billion per quarter. With Google committing $80 billion in capex in 2026, Broadcom is one of the most direct beneficiaries of that spending.

The earnings report tonight is a binary catalyst. If Broadcom raises guidance β€” as it has done consistently over the past six quarters β€” the stock breaks above its all-time high and analysts revise targets upward. The risk is a guidance miss, which would be the first in recent memory. Position sizing matters here: this is a hold-through-earnings situation for existing shareholders, and a buy-on-any-dip opportunity for new entrants if the stock sells off on margin concerns rather than fundamental deterioration.

The buy case in one sentence: Broadcom is the premier custom AI chip company with a direct pipeline to the world's largest hyperscaler capex budgets, and tonight's earnings could be the catalyst that breaks the stock to new all-time highs.

4. AMD (AMD) β€” WATCH (Not a Buy at Current Levels)

Price: $521.54 | 52-week range: $111.01–$527.20 | Forward PE: 59.85 | Analyst consensus target: $479.77 (stock is trading ABOVE the consensus target) | Consensus rating: Strong Buy

AMD's turnaround story is real. The company has taken meaningful share from Intel in the data center CPU market with its EPYC processors, and its MI300X GPU is the only credible alternative to Nvidia's H100 for AI workloads at scale. TD Cowen has a price target of $600, implying 15% upside from current levels β€” and that is the bull case.

The problem is the aggregate consensus. The average analyst target is $479.77 β€” meaning the stock is currently trading above the consensus estimate of the analysts who cover it most closely. When the average Wall Street target implies the stock needs to fall 8% to reach fair value, it is hard to argue for new money in.

There is also the Nvidia threat. RTX Spark, announced at Computex, targets the AI PC and edge computing market that AMD has been positioning itself to capture with its Ryzen AI lineup. If Nvidia successfully commoditizes that segment, AMD's growth trajectory outside the data center becomes murkier.

AMD is a great company at a stretched valuation. Wait for a pullback to the $440–$460 range, which would bring it back in line with the consensus target, before initiating a position.

The watch case in one sentence: AMD is an excellent business, but at $521 it is trading above what the average analyst who covers it believes it is worth β€” wait for a 10–15% pullback before buying.

5. Marvell Technology (MRVL) β€” WATCH (Extraordinary Story, Needs to Settle)

Close price (June 2): $290.79 (+32.52% single session) | Pre-market June 3: $332.56 | 52-week range: $59.53–$291.30 | Forward PE: 64.16 | Previous consensus target: ~$233 | New Stifel target: $321 | Consensus rating: Strong Buy

The Marvell story is the most exciting β€” and the most dangerous β€” on this list. On June 2, MRVL surged 32.52% in a single session after two simultaneous catalysts: a strong earnings report and Jensen Huang's public endorsement at Computex, where he called Marvell "the next trillion-dollar company." At the Evercore TMT Conference on June 3, management raised guidance to $11.5 billion for the current fiscal year and $16.5 billion for the outer year.

The fundamental thesis is powerful: Marvell designs custom AI ASICs for hyperscalers (including reportedly Amazon and Google), networking chips that handle AI traffic inside data centers, and optical interconnect components that become more critical as AI clusters scale. Every dollar spent on AI infrastructure has a Marvell component somewhere in the stack.

But the price action demands respect. The stock has gone from $59.53 to $332.56 (pre-market) in under 12 months β€” a 5x move. The forward PE of 64.16 is high even for a hypergrowth name. The previous analyst consensus was $233; Stifel has rushed out a new target of $321, but much of the Street has not yet updated models. Chasing a stock after a 32% single-day move, with limited fresh analyst coverage, is one of the most common ways traders give back gains.

Watch this name. Let the post-earnings euphoria settle, let analysts update their models, and look for an entry in the $270–$300 range on a pullback. If the thesis is right, there will be another entry point. If there is not β€” if the stock never comes back to that range β€” you will have missed a trade but not made a costly mistake.

The watch case in one sentence: Marvell may be the next trillion-dollar semiconductor company, but chasing a 32% single-day gap with most analyst models still stale is a high-risk entry β€” wait for a pullback and updated price targets before buying.

6. Intel (INTC) β€” AVOID

Price: $107.93 | 52-week range: $18.97–$132.75 | EPS (TTM): –$0.67 (loss-making) | Analyst consensus target: $88.71 (stock is 17.8% ABOVE the consensus target) | Consensus rating: Hold

Intel's comeback narrative is compelling on paper: new CEO Lip-Bu Tan, a restructured foundry business, renewed focus on manufacturing execution, and a stock that has already recovered significantly from its 2024 lows near $18.97. Mizuho has a $128 Neutral price target, suggesting some analysts see limited upside.

The numbers, however, are unambiguous. Intel is still losing money β€” trailing twelve-month EPS of –$0.67. The average analyst who covers Intel has a price target of $88.71, which is 17.8% below where the stock trades today. The consensus rating is Hold, not Buy. These are not the characteristics of a stock you want to own in a sector where capital flows toward growth leaders.

More structurally: Nvidia's RTX Spark announcement is a direct threat to Intel's PC segment, which has been one of the few stabilizing forces in the business. If Nvidia can capture AI PC workloads β€” which its CUDA ecosystem and driver maturity make plausible β€” Intel faces pressure in its last remaining high-volume consumer market.

Intel may be a great comeback story by 2027 or 2028. It is not a buy in June 2026, when the stock is trading 17.8% above analyst consensus, the company is still generating losses, and the competitive threats are intensifying.

The avoid case in one sentence: Intel is loss-making, trading 17.8% above the average analyst target, faces intensifying competition from Nvidia in its core PC market, and carries a Hold consensus β€” there are better places to put capital in this sector.

Summary: The 2026 Semiconductor Trade Sheet

Stock Price (Jun 2–3, 2026) Analyst Target Upside/Downside Rating Verdict
NVDA $222.82 $296.81 +33% Strong Buy BUY
TSM $446.69 $467.84 +4.7% Strong Buy BUY (best risk/reward)
AVGO $481.57 $486.85 +1.1% Strong Buy BUY (earnings catalyst)
AMD $521.54 $479.77 –8% Strong Buy WATCH β€” wait for pullback
MRVL $332.56 (pre-mkt) ~$321 (Stifel, new) –3.5% Strong Buy WATCH β€” wait for settling
INTC $107.93 $88.71 –17.8% Hold AVOID

The Bottom Line

The semiconductor sector is in the middle of a generational growth cycle, not at the end of one. AI infrastructure spending is accelerating, not plateauing β€” and the companies that build, design, and manufacture the chips that power that infrastructure are still in the early innings of monetizing that demand.

That does not mean every semiconductor stock is a buy. Valuation still matters. Price relative to analyst consensus still matters. Competitive positioning still matters. The framework here is simple: buy the companies that are reasonably priced relative to their growth and have strong analyst support. Watch the companies that are great businesses but priced beyond current consensus. Avoid the companies where the numbers β€” losses, above-target prices, Hold ratings β€” are telling you something the narrative is not.

Of the six names covered here, NVDA and TSM offer the clearest buy cases for most investors. AVGO is a buy into tonight's earnings for those who can tolerate binary catalyst risk. AMD and MRVL are exceptional businesses to add to your watchlist and buy on the next significant pullback. Intel is a pass until the turnaround produces actual profits.

The greatest time ever for semiconductors, as Gary Dickerson put it, is not an invitation to buy everything. It is an invitation to be selective β€” and position in the names where the math, the momentum, and the fundamentals all point in the same direction.

Data sourced from StockAnalysis.com as of June 2–3, 2026. This article is for informational and educational purposes only and does not constitute financial advice. Always conduct your own research before making investment decisions.

Official Resources

For further research, the following official sources provide authoritative information on the topics covered in this article.

Sources & Trading Risk Note

This article is for educational purposes only and is not financial advice. Trading involves risk, leveraged products can amplify losses, and market rules or evaluation terms can change. Verify current contract specs, exchange rules, and firm-specific terms before trading.