AI Infrastructure Earnings Signal Robust Demand: Duos Tech Secures 65MW, eMemory Security IP Surges 1,600%

The Quick Take

• Duos Technologies (DUOT) transforms into pure-play AI infrastructure leader with 65 megawatts of contracted capacity and a 2027 revenue target of at least $160 million following Q2 2026 earnings
• eMemory Technology's PUF-based security IP royalties surged over 1,600% year-over-year in Q2 2026, reflecting explosive AI data center adoption of hardware security solutions
• Both companies reported strong contract pipelines tied to AI compute buildout, signaling sustained enterprise demand for AI-grade infrastructure and specialized security IP
• eMemory's expansion into advanced node applications positions it as a key enabler for next-generation AI chip architectures requiring tamper-resistant memory and identity solutions

The simultaneous earnings beats from Duos Technologies and eMemory Technology underscore the broadening infrastructure layer beneath AI's rapid expansion. Duos, once a rail-surveillance technology company, has pivoted decisively to AI data center infrastructure, contracting 65 megawatts of capacity — a scale that positions it among emerging mid-tier AI hosting providers competing for enterprise and hyperscaler overflow demand. The $160 million 2027 revenue framework signals strong execution confidence at a time when AI infrastructure capacity remains a critical strategic bottleneck for cloud providers.

eMemory's 1,600% royalty surge reflects the security-layer demand curve that follows major AI chip adoption cycles. As AI models are deployed across sovereign clouds, edge devices, and enterprise hardware, PUF-based physical unclonable function IP provides the cryptographic identity layer that regulators and enterprise buyers increasingly require. The company's strategic expansion into advanced semiconductor nodes — including sub-3nm processes — ensures compatibility with next-generation AI accelerator designs, creating durable and recurring royalty streams tied to each new chip generation that AI fabless companies bring to market.

Investors should note the differentiated risk profiles these two companies offer within the AI infrastructure investment universe. Duos carries execution risk typical of capital-intensive infrastructure buildouts, where delays in power procurement and construction timelines can compress projected returns, while eMemory's royalty model provides higher-margin, recurring revenue with substantially lower capex intensity. Together, these results reinforce the thesis that AI infrastructure spending extends well beyond GPU procurement to encompass compute hosting, power infrastructure, and embedded security IP — sectors that may exhibit more durable earnings trajectories than headline AI software plays and merit dedicated portfolio consideration.

Synthesized from 2 sources.