Compare GPU clouds / Firmus
Firmus
Asia-Pacific AI cloud and AI factory builder (formerly Sustainable Metal Cloud)
Large clusters
Firmus operates GPU cloud capacity in Singapore (as SMC by Firmus) and Australia, with Malaysian and Indonesian sites under development, and offers H200, H100, A100 and L40S on demand or reserved, plus GB300 NVL72. It reports more than 900 MW of contracted capacity across seven AI factories in four countries. It is rated Silver in SemiAnalysis ClusterMAX 3.0, which notes Kubernetes and Slurm multi-node NVLink setups.
Best for: GPU training and inference capacity in Asia-Pacific.
ClusterMAX 3.0: Silver
- – Publishes prices
- – Serverless inference
- – AMD GPUs
- – European data centers
Published prices
Where we recommend them
- Best for Asia-Pacific, our pick.
Worth knowing before you sign
- No published GPU prices; access is by enquiry or reservation.
- Its business increasingly centres on large contracted AI factories (OpenAI anchor, more than 900 MW contracted) rather than self-serve cloud.
- No security certifications are stated on the pages checked.
Sources
- Firmus
- Firmus AI Cloud compute
- Firmus about
- Firmus: US$2B equity round
- Firmus: 900 MW and OpenAI anchor
- SMC by Firmus
- SemiAnalysis ClusterMAX 3.0 ratings
- SemiAnalysis ClusterMAX 3.0 report
Work at Firmus? Send corrections with a source to editors@bestneoclouds.com. We fix verified errors within five business days.
Compare with
- CoreWeave GPU cloud built around large InfiniBand training clusters (NASDAQ: CRWV)
- Nebius Full-stack AI cloud with published on-demand GPU prices (NASDAQ: NBIS)
- Lambda GPU cloud with self-serve instances and 1-Click Clusters
- Crusoe AI cloud with self-serve NVIDIA and AMD GPUs on its own data center campuses
- Nscale European AI cloud selling GB300 and B200 capacity (S-1 filed for NYSE: NSCL)