01
Bare Metal & GPU Compute
Deploying and managing automated bare metal servers and dedicated GPU clusters, including plan/spec selection, location choice, and the hourly vs. monthly vs. yearly billing options surfaced in pricing.
“Deploy and manage fully automated bare metal servers globally in under 5 seconds.” www.latitude.sh
Mapped capabilities
4 capabilities
Plan and spec selection
Matching CPU/RAM/NVMe/network specs (e.g. rs4.metal.xlarge, c4.metal.large) to a stated workload.
GPU cluster configuration
Reasoning over GPU plans, VRAM per GPU vs. total, and interconnect details (e.g. g4.b300.large, g3.h100.small).
Location and global footprint
Choosing among the 25 locations and explaining per-plan location availability.
Billing mode trade-offs
Hourly vs. monthly vs. paid-yearly pricing and included bandwidth (20 TB free out / unlimited in).




