Explore the foundational technologies and hardware driving the artificial intelligence boom. From next-generation data centers and custom silicon to advanced networking and energy solutions, discover the deep tech investments building the physical and digital layers that power the AI era.
Positron AI closed an $875M Series C at a $5B valuation on a contrarian bet: skip high-bandwidth memory, put cheap LPDDR5X next to a custom datapath, and serve very long contexts from a single node. The chip that has to prove it is not in production yet.
Gradient-backed pre-seed for Nebula, a hierarchical state layer that lets long-horizon agents query current enterprise workflows instead of rebuilding context every turn.
RHAELM says co-location at JERA's Chiba Thermal Power Station can deliver 400 MW of AI compute years ahead of a grid-connected schedule. Here is what the memorandum actually commits, and what it leaves to 2028.
Electrical HBM hits a hard limit on capacity and bandwidth. Volantis is betting a photonic fabric with integrated micro-VCSELs can dissolve that trade-off.
On 29 September 2026 Reco added $55M to reach $140M in total while Rig Security left stealth with $12M, both betting that non-human identity governance is a market. CSA research puts the underlying gap at 20% revocation maturity and 1 in 20 identities holding full admin.
Efficient Computer signed agreements for more than $97M at a $650M valuation to move a validated edge architecture into the data centre. Electron E1 runs on 1,000+ physical AI robots at up to 1 TOPS per watt. The data centre chip has no name and no tape-out date.
Token prices fell roughly 600-fold since June 2020 and enterprise cost per workload dropped 67% year over year, yet inference spend is still forecast to triple to $885B by 2030. Four delivery models now compete for the same workload: hyperscaler cloud, neocloud, neutral exchange, on-premises.
SiMa.ai raised $150M at a $1.45B valuation to run AI on devices rather than the cloud. Fidelity and Amplify co-led a round that brings total funding to $500M, and sets a 2028 deadline for the edge-inference thesis.