In this article, we take a look at the 10 Best Performing AI Stocks Over the Last 3 Years.
The strongest AI-linked stock returns of the past three years have often followed the physical buildout of computing capacity rather than the broad software narrative alone. Chips, memory, optical links, storage, power systems, cooling equipment, and rack integration have all become important parts of a capital-intensive supply chain. Stanford HAI’s 2026 AI Index puts global corporate AI investment at $581.7 billion in 2025, up 130% from 2024, while private AI investment reached $344.7 billion. The scale of deployment is also visible in electricity use. The International Energy Agency says data centers consumed about 485 terawatt-hours in 2025, with demand rising 17%. Electricity use at AI-focused facilities increased 50%.
Spending remains elevated. Reuters reported that JPMorgan expected five large technology companies to invest roughly $730 billion in 2026. That supports demand across the infrastructure chain, but it also raises the standard for execution. Power availability, component supply, financing costs, and the pace of customer monetization may determine which businesses retain pricing power. PIMCO has estimated that capital spending could absorb 94% of major hyperscalers’ operating cash flow over the next two years, compared with 40% in 2023. The industry backdrop is therefore substantial, though high valuations and the possibility of overbuilding still warrant some restraint.
Source: unsplash
Methodology
We screened U.S.-listed companies with material exposure to the AI value chain and reviewed their full three-year trading history through the July 9 close. From these, we selected 10 companies with the highest 3-year annualized returns and ranked them accordingly.
Why are we interested in the stocks that hedge funds pile into? The reason is simple: our research has shown that we can outperform the market by imitating the top stock picks of the best hedge funds. Insider Monkey’s quarterly newsletter’s strategy selects 14 small-cap and large-cap stocks every quarter and has returned 599.2% since May 2014, beating its benchmark by 372 percentage points (see more details here).
10. Ciena Corporation (NYSE:CIEN)
Ciena Corporation (NYSE:CIEN) is one of the best performing AI stocks over the last 3 years, with a 3Y CAGR of 120.4%. On June 8, UBS reiterated its Hold rating on the stock and raised its price target to $508 from $285. The action followed Ciena’s fiscal second-quarter report, where revenue rose 40% year over year to $1.57 billion, and adjusted earnings came in ahead of FactSet estimates, MarketWatch reported. UBS analyst David Vogt’s caution centered on expectations rather than the headline numbers.
He wrote that the market had been pricing in a “more material beat and raise” than Ciena delivered. Ciena still raised its fiscal 2026 revenue outlook to $6.2 billion to $6.4 billion and guided fiscal third-quarter revenue above Wall Street estimates. The gap between strong results and an even stronger setup explains the neutral stance: Ciena remained tied to AI-driven optical networking demand, but the stock had already climbed sharply into the report. Ciena also said its high-speed connectivity strategy across wide-area networks and data centers was aligned with multi-year AI demand.
Ciena Corporation (NYSE:CIEN) provides networking systems, services, and software used by communications service providers, cloud operators, governments, and enterprises to move, manage, and optimize high-bandwidth network traffic.
9. Vertiv Holdings Co (NYSE:VRT)
Vertiv Holdings Co (NYSE:VRT) is one of the best performing AI stocks over the last 3 years, with a 3Y CAGR of 134%. On June 12, Vertiv completed its acquisition of ThermoKey, an Italian provider of heat-rejection and heat-exchange equipment. The transaction adds dry coolers, heat exchangers, and systems compatible with low-global-warming-potential and natural refrigerants. It also expands manufacturing capacity in Europe, the Middle East, and Africa. That is relevant to AI infrastructure because rising rack densities are making cooling design a facility-level constraint, not merely an accessory to server deployment.
Vertiv has already used ThermoKey technology in selected products, which may reduce integration risk. The combination also extends Vertiv’s coverage across the thermal chain, allowing it to address heat removal beyond the immediate rack. The investor question is whether the acquired capacity can be converted into profitable growth without weakening execution. The deal appears strategically aligned with demand for denser computing, but returns will still depend on project timing, pricing and disciplined integration.
Vertiv Holdings Co (NYSE:VRT) supplies critical digital infrastructure, including power management, thermal management, racks, monitoring systems, and lifecycle services for data centers, communications networks, and commercial and industrial facilities.
8. Lumentum Holdings Inc. (NASDAQ:LITE)
Lumentum Holdings Inc. (NASDAQ:LITE) is one of the best performing AI stocks over the last 3 years, with a 3Y CAGR of 143.31%. On June 11, JPMorgan analyst Samik Chatterjee reiterated an Overweight rating and maintained a $1,130 price target, arguing that the recent pullback in optical-networking stocks had created a buying opportunity.
Chatterjee said Lumentum shares had fallen about 15% from their early-June high as investors worried about limited summer catalysts and possible delays in co-packaged optics adoption. He pushed back on those concerns, citing supply-chain checks indicating that Nvidia’s large-scale CPO rollout remained on track and may be ahead of schedule. JPMorgan also pointed to growing interest from cloud-service providers and other customers, which could broaden demand beyond Nvidia.
Co-packaged optics places optical engines close to switch chips to increase bandwidth while reducing power consumption in large AI systems. JPMorgan viewed Lumentum’s valuation of roughly 25 times estimated 2028 earnings as reasonable, given projected annual earnings growth above 40%. The thesis still depends on deployment schedules, customer concentration and Lumentum’s ability to convert product demand into sustained shipments and earnings growth.
Lumentum Holdings Inc. (NASDAQ:LITE) develops optical and photonic technologies for AI and cloud networking, telecommunications, industrial manufacturing and sensing, including lasers, components, modules and optical subsystems.
7. Applied Optoelectronics, Inc. (NASDAQ:AAOI)
Applied Optoelectronics, Inc. (NASDAQ:AAOI) is one of the best performing AI stocks over the last 3 years, with a 3Y CAGR of 144.7%.
On June 21, Rosenblatt Securities analyst Michael Genovese reiterated a Buy rating and maintained a $220 price target. The firm’s documented thesis has centered on Amazon-related 800G revenue, potential Oracle qualifications and a product mix spanning 100G, 400G, 800G and 1.6T connectivity. That view is supported by more than $324 million of reported orders for 800G and 1.6T products tied to hyperscale demand. That backlog makes execution the focus. Orders and customer qualifications must convert into shipments, while higher volumes need to improve profitability and cash flow. AAOI’s vertically integrated manufacturing model can help with cost and supply control, yet rapid capacity expansion creates its own operational burden.
Customer concentration is another important consideration, as a delayed qualification or a changed deployment schedule can materially affect results. Rosenblatt’s reiteration indicates confidence in the optical ramp, while the order base provides a measurable benchmark for subsequent delivery.
Applied Optoelectronics, Inc. (NASDAQ:AAOI) designs and manufactures optical networking products, including lasers, transceivers and related equipment for internet data centers, cable television networks, telecommunications and fiber-access markets.
6. Seagate Technology Holdings plc (NASDAQ:STX)
Seagate Technology Holdings plc (NASDAQ:STX) is one of the best performing AI stocks over the last 3 years, with a 3Y CAGR of 151%. On July 1, Bank of America analyst Wamsi Mohan reiterated a Buy rating and raised his price target to $1,150 from $1,000. Mohan’s thesis pointed to improving fundamentals, sustained cloud demand and greater discipline in adding hard-drive capacity. He also highlighted build-to-order contracts that support measured price increases, along with the transition to higher-capacity heat-assisted magnetic recording drives.
Those factors could improve revenue visibility, margins, and free cash flow if customer demand remains firm. The AI link is primarily storage economics: training and inference create large datasets that cloud operators must retain at scale, where high-capacity hard drives remain cost-effective. The principal risk is cyclical. Strong pricing can encourage future supply, while customers can pause purchases after capacity additions. The current analyst case rests on disciplined industry behavior and a durable data-center demand cycle, not demand alone.
Seagate Technology Holdings plc (NASDAQ:STX) develops mass-capacity data-storage products and systems, including hard disk drives and enterprise storage used by cloud, edge, consumer, and AI data-center customers.
While we acknowledge the potential of STX to grow, our conviction lies in the belief that some AI stocks hold greater promise for delivering higher returns and have limited downside risk. If you are looking for an AI stock that is more promising than STX and that has 100x upside potential, check out our report about the cheapest AI stock.
Click to continue reading and see the 5 Best Performing AI Stocks Over the Last 3 Years.
Disclosure: None. Follow Insider Monkey on Google News.
