Menlo Cloud

Menlo Cloud is our private cloud for robotics training and development, located across 2 availability zones in Singapore.

• Specialized robotics infrastructure: We built Menlo Cloud because humanoid robotics requires specialized hardware that hyperscaler clouds don’t offer: ARM-based edge compute boards for firmware testing, robot-specific acceleration, and hardware-in-the-loop validation at scale.

• Predictable access and unit economics: We eliminate surprise pricing from hyperscalers while ensuring we have a guaranteed baseline access to compute, even during times of high demand.

Infrastructure

Built on OpenStack with a diverse GPU fleet:

• Nvidia, AMD, and Intel GPUs for cross-platform development and CI/CD
• Colocated across 2 availability zones in Singapore for high redundancy
• Custom GPU servers for non-traditional hardware builds (e.g. embedded ARM boards)

History

Menlo has always run its own infrastructure. We started out small:

Menlo’s original “GPU cloud” in May 2023 — two 4090s in our founder’s spare bedroom

• Started with two RTX 4090s running in our founder’s spare bedroom
• Grew to multiple machines running in our Infra Lead’s spare bedroom
• Evolved into a full internal cloud with strong uptime and reliability guarantees

GPU Servers

Eventually, as Menlo’s hardware footprint grew, we built custom GPU servers to serve our needs: