Made obsolete by the rise of software, old IT glories like Dell, Cisco, Nokia and Texas Instruments are experiencing a strong comeback on Wall Street. Their know-how in IT infrastructure has once again become strategic with the explosion of AI.
After clothing, cinema and music, retromania is winning over the stock market. The hottest stocks on Wall Street are no longer just those of the Magnificent Seven and the latest AI start-ups that have entered the public markets, but of the gray beards of IT whose names evoke more of the 1990s than the future, like DellTexas Instruments, Lenovo and Nokia.
Once out of date in the face of booming software startups, these companies, specializing in IT infrastructures, are today experiencing a strong comeback driven by AI, which requires precisely the construction of a large amount of IT infrastructure to operate.
Infrastructure providers hit the jackpot
Dell’s action, including server sales are boosted by the construction of AI data centers, is now selling at a price four times higher than in February. The company has in fact been able to position itself as a key player in the assembly of AI servers, which require interconnecting thousands of racks and making them operate on a cluster scale. A field in which Dell excels thanks to forty years of experience in the manufacturing of servers, to which will be added, in 2024, a strategic partnership with Nvidia around the vision of “AI factories” promoted by Jensen Huang. Rather than simply selling components, Dell has focused on providing turnkey infrastructures (the servers are optimized for Nvidia GPUs, with high-performance storage and network systems), pre-integrated and quickly deployable. This is exactly what hyperscalers are looking for who want to quickly scale their infrastructures without having to set up specialized engineering teams each time.
The action of his counterpart Cisco has jumped 50% since the beginning of April, also driven by record orders from hyperscalers. An old veteran of American IT, Cisco has experienced several technological failures, including its purchase of the professional videoconferencing system Webex, a pioneer in the field which never managed to establish itself and was supplanted by Zoom and others. But the group’s expertise in telecommunications networks gave it a card to play that it was able to exploit around AI. This in fact requires networks internal to data centers that are infinitely faster and denser: the GPU clusters mentioned above constantly exchange colossal volumes of data in real time, which creates a massive demand for switching, routing and high-performance network infrastructures, an area on which Cisco has built its expertise, and which now allows it to benefit fully from orders coming from hyperscalers.
Let’s quote Chinese again Lenovo. Mainly known for its personal computers, the group also offers servers dedicated to AI, liquid cooling solutions to manage the thermal increase in hardware linked to high-performance computing, as well as storage and virtualization tools. It is therefore also benefiting from the boom in the construction of AI data centers and is experiencing excellent stock market performance.
How the veterans of telephony came back strong thanks to AI
After very lackluster years, Nokia’s stock is one of the most successful in Europe: the former telephony giant, made obsolete by the smartphone era, markets optical equipment of which AI data centers are particularly fond, the workloads linked to this technology requiring ever faster and more efficient data transfers between computing clusters. However, optical fiber allows data transfers via light pulses, faster than the electrical signals traveling on copper cables, which are used in traditional data centers.
Blackberry, another name linked to telephony from the 2000s, has also seen its share price triple in recent months, riding in particular on the performance of its embedded software division, QNX. Although the BlackBerry brand has disappeared from the smartphone market, its software remains integrated into more than 275 million vehicles around the world today. QNX is particularly used for critical functions related to safety and reliability (advanced driver assistance systems), lane keeping, collision alert and software-defined vehicle architectures).
Beyond automotive, QNX is also deployed in sectors such as medical, robotics, industrial and AI, leveraging its expertise gained in automotive software around resilience, cybersecurity and software for critical systems. “We are currently expanding into adjacent markets that have a profile similar to that of automotive. In AI, for example, we work closely with Nvidia to guarantee a high degree of functional security. We are not really involved in the AI programming itself, but we ensure that their AI frameworks work correctly on our platform. Customers can then use CUDA or any other tool they offer, at the application level”, details John Wall, president of QNX.
A bath of youth for semiconductor veterans
The inference boom also offers opportunities to semiconductor specialists who missed the first wave of AI. This is the case of Texas Instruments, whose calculators may bring back middle school memories to some. Based in Dallas, this company is primarily a specialist in computer chips, particularly those used in telecommunications equipment. Although it initially completely missed the AI shift, the company is selling more and more chips to power servers that require ever higher power density. Its data center division now generates more than a billion dollars in annual revenue, and grew by more than 60% last year. Its share price is currently around 300 dollars, compared to 200 at the beginning of March.
Let us also mention Micron, a specialist in memory chips, subject to a very high demand with the inference boom. The company crossed the trillion-dollar market capitalization mark for the first time on May 26, 2026, almost fifty years after its creation. Or Intel, in great difficulty just a year ago, and which has relaunched itself in the race thanks in particular to a succession of strategic partnerships around AI.
These companies benefit from a strong business model and strong cash flow
“These companies have been transformed by the AI era, and the robust businesses they were quietly running now offer huge opportunities for productivity and expansion. They know how to build the hardware, they have the skills, the supply chains, and that puts them in an ideal position to take advantage of the expansion of AI,” said Michael McCarthy, market strategist at Moomoo Australia, a brokerage platform, in a show he hosts on YouTube.
For him, “thanks to their established activities and their recurring cash flows, these companies are ideally positioned, especially at the dawn of the next phase of AI expansion, where many companies will need to raise capital: those which already have significant cash flows have a considerable advantage. However, in most cases, these old tech stocks display solid balance sheets. They therefore have cash. They also benefit fully from the “AI supplier effect full stack”: they can indeed expand vertically within their sectors, progressively and without jeopardizing their core business.”
