Amp

Hydra Host, founded in 2021 in Miami, USA, matches idle GPU capacity in data centers with enterprises and developers, providing a compute scheduling and monetization operating system to address the supply-demand mismatch in computing resources for AI companies and developers.

Dynamia.ai, founded in China in 2024, focuses on heterogeneous computing power, the HAMi open-source ecosystem, and vGPU scheduling. The company leads the HAMi open-source community and drives deployment of heterogeneous computing scheduling platforms.

Leinao (Hefei Zhongke Leinao Intelligent Technology Co., Ltd.), founded in September 2017, was spun out of the National Engineering Laboratory for Brain-Inspired Intelligence Technology and Application and USTC's Future Network Laboratory, and is headquartered in China Sound Valley, Hefei High-Tech Zone. It was the first Chinese company built around brain-inspired intelligence technology. Its current focus is a "power-compute co-optimized token factory" spanning heterogeneous compute scheduling, token inference optimization and energy management: it operates and aggregates more than 5,000 PFLOPS of compute and provides intelligent operations and maintenance for over a hundred electrical substations, using a "1+3+N" product framework to address the mismatch between green power supply and AI compute demand. Chairman and CEO Liu Haifeng is a USTC graduate and a joint USTC-Microsoft Research Asia PhD.

Render Services, founded in 2018 in San Francisco, USA, provides a cloud computing platform that helps developers build, deploy, and scale applications without managing infrastructure.

Aria Networks, founded in 2025 in Palo Alto, California, USA, builds networking infrastructure for AI data centers.

Panthalassa was founded in 2016 in Portland, Oregon, USA, building autonomous offshore wave energy nodes that convert ocean wave energy into electricity and support offshore AI computing infrastructure.