Semiconductor startup Poseidon has successfully closed a $230 million Series B funding round. According to sources familiar with the matter, the company intends to use the capital to accelerate the deployment of its high-speed memory chips, which are critical core components for artificial intelligence computing processors.
Insiders revealed that investors in this round include the Qatar Investment Authority (QIA), the nation's sovereign wealth fund, which has been increasingly focusing its attention on artificial intelligence infrastructure investments.
This Series B financing for the Reno-based startup comes at a time when hyperscale cloud providers and AI companies are actively working to reduce their reliance on the long-time industry leader, NVIDIA. This group includes OpenAI—despite being one of NVIDIA's largest and most significant clients, reports indicate the company has been dissatisfied with some of NVIDIA's latest AI chips and has been seeking alternatives since last year.
Simultaneously, Qatar is accelerating the development of large-scale domestic AI infrastructure through QIA, a priority that was repeatedly emphasized during the Qatar Web Summit held in Doha this week. Multiple sources told TechCrunch that Qatar views computing power as essential for maintaining competitiveness on the global economic stage and is committed to establishing itself as a leading AI services hub in the Middle East, which in turn is driving investment interest in startups like Poseidon.
This strategy is already being implemented through several major collaborations, including a $20 billion AI infrastructure joint venture announced this September in partnership with Brookfield Asset Management.
With the completion of this funding round, Poseidon, which was founded just three years ago, has now raised a total of slightly over $300 million. The company previously secured $75 million in funding last year from investors including Valor Equity Partners, Atreides Capital, Draper Fisher Jurvetson Growth Fund, Froom Ventures, and the Resilience Capital Reserve Fund.
Poseidon claims that its first chip, named Atlas, which is mass-produced in Arizona, delivers performance comparable to NVIDIA's H100 graphics processor while consuming less than one-third of the power. As enterprises shift from building large-scale models to deploying them at scale, demand for inference hardware is surging. Poseidon's focus on the inference segment—providing the computing power needed to run AI models in practical applications, rather than training large language models—aligns well with current market needs.
Sources indicate that beyond its advantages in memory, Poseidon's chips also demonstrate strong performance in high-frequency computing and video processing scenarios.
Comments