Category Supercomputing

Inertia says it has cut fusion fuel pellet production from days to minutes

Inertia says its fusion fuel pellets now take hours, not days, easing tritium costs and bringing its commercial plant plan closer.

Nvidia’s compute financing push could turn GPUs into collateral, attract Wall Street capital, and reshape AI infrastructure lending.

Relativity Networks raised $22M to expand hollow-core fiber, aiming to cut data-center latency as AI campuses spread out.

Meta’s Glimmer release revives the AI for everyone debate as Zuckerberg promotes openness while keeping some models behind APIs.

French startup Kog is betting GPU inference software can unlock major speed gains on standard datacenter chips and attract enterprise customers.

Situational Awareness pours $400M into a chip startup as the AI hedge fund cuts public holdings and doubles down on infrastructure.

Mirendil signed a Google Cloud deal worth over $100M to power self-improving AI research, highlighting the fight for compute.

AMD’s AI data center business more than doubled to $6.7 billion, while gaming revenue fell 31% in a mixed Q2 report.

Texas orders data center audits as ERCOT’s queue hits 474 GW, raising questions about power, water, and grid reliability.

Runware launches a modular data center pod for faster, cheaper AI inference, betting portability can beat traditional buildouts.