4 private companies helping power the infrastructure of AI
Artificial intelligence has become one of the most powerful investment themes in private markets, but the story extends far beyond the popular consumer-facing applications of the day. Behind every AI breakthrough sits a rapidly expanding infrastructure stack encompassing compute, networking and storage architectures designed to support increasingly demanding workloads.
As enterprises race to deploy AI at scale, a new generation of private companies is building the foundational technologies that aim to power the future of artificial intelligence. This week, we’re featuring four such private companies below. Notably, all four have seller asks on the Forge Marketplace as of June 4, 2026, creating potential opportunities for accredited investors seeking exposure to critical AI infrastructure providers.
PsiQuantum, pursuing quantum computing
Palo Alto, California-based PsiQuantum is pursuing an ambitious goal of building a fault-tolerant, utility scale quantum computer. In simple terms, quantum computers utilize qubits, which can represent multiple states of information, allowing them to solve highly complex simulation problems far faster than today’s powerful supercomputers.
In September 2025, PsiQuantum raised $1 billion in new funding at a valuation of $7 billion, with a notable investor being AI chipmaker NVIDIA.1 In March, PsiQuantum announced it had integrated NVIDIA’s CUDA-Q platform into Construct, its software suite for developing quantum applications. The partnership allows researchers and developers to use NVIDIA’s GPUs to simulate large-scale quantum algorithms and stress-test at a scale previously impractical.2
The collaboration with NVIDIA highlights a growing convergence of AI and quantum computing technology. As AI models grow more complex and quantum systems become more capable, many industry leaders envision a future where AI and quantum computing work together within hybrid computing environments.3
Founded in 2015, PsiQuantum’s Forge Price™ is $30.22 as of June 5, implying a valuation of $5.14 billion. Other VC investors of the private company include Ribbit Capital, Redpoint Ventures, BlackRock and Temasek Holdings.
Lightmatter, developing photonic computing
Boston, Massachusetts-based Lightmatter is developing photonic computing and networking technologies designed to address moving massive amounts of data efficiently between processors, a pressing challenge of today’s AI systems. Lightmatter’s photonic technology uses light to pass information, which moves faster than electricity through metal wires found in traditional compute systems.
On June 3, Lightmatter announced that it has joined AI chipmaker NVIDIA’s NVLink Fusion ecosystem, a program designed to help partners build the next-generation AI infrastructure that integrates seamlessly with NVIDIA’s industry-leading AI platform. This in turn will bring Lightmatter’s photonic interconnect technology into NVIDIA-powered AI systems.4
For Lightmatter, the partnership with NVIDIA represents a significant validation of its technology. As AI models continue to scale, the process of transferring data via electricity, which consumes more energy, is increasingly being seen as a blocker to the broader deployment of AI.5 Lightmatter is positioned to potentially benefit from this growing need for a more efficient alternative.
Founded in 2017, Lightmatter’s Forge Price™ is $89.00 as of June 5, implying a valuation of $4.88 billion. The firm’s notable investors include Sequoia Capital, Bossa Invest, Calm Ventures and Eastward Capital Partners.
Tenstorrent, building computers for AI
Headquartered in Toronto, Canada, Tenstorrent designs specialized computer chips that help artificial intelligence systems process information more efficiently. Just as graphics processing units (GPUs) became essential for training and running modern AI models, Tenstorrent is developing alternative AI processors designed to handle the massive calculations required by generative AI and machine learning applications.
In 2024, the private company received a $700 million investment in a funding round led by South Korea’s AFW Partners and Samsung Securities, with participation from Bezos Expeditions and LG Electronics. At the time, the capital was to be used to build out Tenstorrent’s engineering team and optimize its supply chain to further develop its technology and take on the AI technology giant NVIDIA.6
More recently, it was reported that Qualcomm (Nasdaq: WCOM) had expressed interest in acquiring Tenstorrent. Should the merger occur, Qualcomm would combine its mobile and wireless services with Tenstorrent’s AI accelerators and potentially position itself as a leader within the broader semiconductor industry.7
Founded in 2016, Tenstorrent’s price per share is $80.85 as of its latest funding round in November of 2025.
VAST Data, building an AI operating system
New York-based VAST Data has developed a platform that combines storage, database and data management capabilities into a unified architecture designed specifically for large-scale AI environments. In other words, VAST Data is designed to help solve the challenge of delivering enormous amounts of data required by AI systems.
Founded in 2016, the firm closed a $1 billion funding round in April, which put it at a valuation of $30 billion. The round was co-led by Drive Capital and Access Industries with participation from Fidelity and NVIDIA, among others. According to the company, the capital will be put towards positioning itself as the premier AI infrastructure leader, fueling its growth and gaining strategic partnerships across the AI industry.8
More recently, VAST Data announced a partnership with Mistral AI and Mistral Compute to help power a new generation of NVIDIA-accelerated AI factories across Europe.9 The collaboration brings together NVIDIA’s computing power, Mistral’s AI models and VAST Data’s efficiency, representing a clear example of the convergence of technology firms to optimize AI and its technology.
Investor Takeaway
While much of the public conversation around artificial intelligence focuses on models and applications, the infrastructure layer continues to attract substantial capital and strategic interest. Companies such as PsiQuantum, Lightmatter, Tenstorrent and VAST Data are building technologies that could enable the next decade of AI innovation.
For accredited investors seeking exposure to this rapidly evolving sector, the presence of seller asks for all four companies on the Forge Marketplace showcases potential opportunities to access some of the private companies helping build the foundation of the AI economy.
View more companies within the computing hardware sector.
