Semiconductor startups are commanding higher median valuations than AI firms amid rising chip and computing infrastructure demand, while AI-driven supply constraints threaten to raise smartphone manufacturers’ costs, writes JUSTICE OKAMGBA
S
emiconductor startups are commanding higher valuations than artificial intelligence companies among newly minted billion-dollar private businesses in 2026, reflecting growing investor appetite for the hardware and infrastructure powering the global AI boom.
New research conducted by BestBrokers, showed that semiconductor startups that attained unicorn status this year had a median valuation of $2.5bn, the highest among 12 industries analysed. A unicorn is a privately held startup valued at $1bn or more, a milestone that reflects its perceived worth among investors. Semiconductor companies accounted for only 16 of the 256 new unicorns recorded during the period.
Semiconductors are materials used to manufacture electronic chips that control how electricity flows through devices. They are essential components of modern electronics, including smartphones, computers and data centre equipment.
The median valuation of semiconductor startups was $1bn higher than the $1.5bn recorded by AI startups, which produced the largest number of new unicorns, with 63 companies crossing the billion-dollar threshold. Cryptocurrency and blockchain companies ranked second, with a median valuation of $1.95bn, while robotics followed at $1.45bn.
The findings highlight how investment in artificial intelligence is extending beyond companies developing models and applications to businesses supplying the semiconductors, networking equipment and computing infrastructure required to operate increasingly sophisticated AI systems.
Despite their lower median valuation, AI startups accounted for the largest number of new billion-dollar companies and a substantial share of their combined value. The report, which analysed the Crunchbase Unicorn Board alongside data from PitchBook and other sources on private-company funding rounds and valuations, found that the 63 new AI unicorns were collectively valued at $236.1bn, representing about 42 per cent of the combined value of all startups that attained unicorn status in 2026.
Chinese artificial intelligence company DeepSeek accounted for $91bn of the aggregate valuation, making it the most valuable newly minted unicorn identified in the report.
However, semiconductor companies attracted higher median valuations despite their smaller numbers, with optical networking startup Lumilens leading the sector at $5.5bn. Chipmaking equipment company Source Foundry and AI inference chipmaker Positron followed, each valued at $5bn.
The higher valuations reflect growing demand for advanced computing components as technology companies expand their artificial intelligence infrastructure, increasing pressure on the semiconductor supply chain.
Global semiconductor sales reached $517.7bn in the third quarter of 2026, up 24.1 per cent from the preceding quarter, according t figures from the Semiconductor Industry Association cited in the report.
The research identified shortages of advanced chips, limited capacity to transfer data between processors, and rising electricity requirements for data centres as emerging constraints on AI infrastructure expansion.
Commenting on the findings, Financial Analyst at BestBrokers, Stanislava Savisheva said artificial intelligence remained the biggest driver of new billion-dollar private companies, but semiconductor businesses were benefiting from the infrastructure requirements of the technology.
“AI is the undisputed engine of the startup world right now. It has minted more unicorns than any other industry this year; it produced DeepSeek, which is the most valuable newly minted startup in 2026, as well as the two most valuable private companies on the planet, Anthropic and OpenAI,” Savisheva stated.
She noted that the growing demand for computing power was strengthening the investment case for semiconductor companies.
- Police deny declaring man wanted for impregnating eight women in Cross River
- Huawei warns of AI-driven cyber risks
- Nvidia rules out Huang’s return as CEO, reports revenue surge
“But you can’t run chatbots and agents without chips. Sixteen semiconductor startups have become unicorns so far this year, and the typical one is valued at around $1bn more than the typical new AI unicorn. With global chip sales hitting $517.7bn in a single quarter, it’s clear where a lot of the AI money ends up,” she added.
Savisheva said the relationship between AI developers and semiconductor suppliers was increasingly shaped by constraints in computing infrastructure, data movement and electricity supply.
“AI and semiconductors have formed a sort of equilibrium, but as AI grows, it has begun to run into hard limits: there are only so many advanced chips, only so much capacity to move data between them, and only so much electricity to keep data centres running. Every one of those bottlenecks is a business opportunity. Solve one, and you’ve found the goose that lays the golden eggs,” she said.
Beyond semiconductors and AI, cryptocurrency and blockchain startups recorded the second-highest median valuation of $1.95bn, although only five companies in the sector became unicorns during the period.
Three of them, Digital Asset, TRM Labs and Rain, were valued between $1.95bn and $2bn, placing the sector’s typical new unicorn about 30 per cent above its AI counterpart.
Robotics ranked fourth, with a median valuation of $1.45bn across 30 new unicorns. The companies had a combined valuation of $58.9bn, the second-highest aggregate value among the industries analysed.
China accounted for 18 of the new robotics unicorns, led by XPENG Robotics, which had a private valuation of $6.3bn.
Defence technology ranked fifth, with a median valuation of $1.3bn across 16 new unicorns. Germany’s STARK, valued at $4bn, and US-based Mach Industries, valued at $3.7bn, were the sector’s most valuable new entrants.
Fintech followed with a median valuation of $1.25bn, led by insurance startup Corgi, which was valued at $4bn.
Cybersecurity ranked seventh at $1.2bn, while energy and climate technology and cloud and infrastructure recorded median valuations of $1.15bn each.
Enterprise software and aerospace and space technology followed, both with median valuations of $1.1bn, while health technology and biotechnology ranked last at $1bn.
The health and biotechnology sector recorded 21 new unicorns, with 11 valued at the minimum $1bn threshold required to qualify for the category.
Beyond the investment opportunities for semiconductor companies, the competition for computing components is also beginning to affect manufacturers of consumer electronics, particularly smartphones and personal computers.
Rising demand for high-bandwidth memory used in artificial intelligence data centres has tightened supplies of conventional memory chips used in consumer devices, pushing up component costs and forcing manufacturers to reconsider production and pricing strategies.
IDC warned that the shift towards AI-related memory production could constrain supplies of components used in consumer electronics, while Counterpoint Research projected a decline in global smartphone shipments in 2026 as manufacturers grappled with rising memory costs.
The developments highlight the wider implications of the artificial intelligence boom. While AI is producing more billion-dollar startups than any other industry, the investment opportunity is increasingly spreading to businesses supplying the chips, networking equipment and computing infrastructure needed to sustain its expansion. At the same time, competition for these components could increase costs for consumer electronics manufacturers, linking the growth of AI infrastructure to the production and pricing of everyday devices.

No comments yet. Start the conversation.