Company updates on funding, infrastructure, and the buildout of the world's fastest inference cloud.
Debt facility
$400M
Upper90 and other GPU financiers are moving beyond training GPUs to back inference-native ASIC silicon. This facility lets us finance the buildout of inference capacity at scale.
We secured a $400M debt facility led by Upper90 -- the first GPU financiers now betting on inference-native ASIC silicon.
We raised $15M to build the world's fastest neocloud, powered by ASIC chips purpose-built for AI inference.