Firmus’ Sustainable Metal Cloud ranks among world’s top GPU clouds, outperforming global giants

SINGAPORE - Firmus Technologies, the AI infrastructure company behind the high-performance Sustainable Metal Cloud (SMC), has been recognised by analyst firm SemiAnalysis as one of the top-performing GPU cloud platforms globally, ranking alongside established hyperscalers AWS and above Google Cloud.

In the first-ever ClusterMAX™ GPU Cloud Ratings, Firmus’ Sustainable Metal Cloud was named one of only three GPU providers worldwide to have correctly implemented InfiniBand SHARP in-network reduction. This technically complex performance enhancement dramatically boosts throughput for AI workloads.

“It’s fantastic to see Sustainable Metal Cloud put to the test and come out ahead of the global pack,” said Tim Rosenfield, Co-CEO of Firmus. “The industry is evolving fast - and it’s great to be recognised for staying ahead of the curve. We’ve invested heavily in the infrastructure that matters: scalable, liquid, energy-efficient systems that deliver real outcomes for customers.”

The high-performance reputation of the Sustainable Metal Cloud is built on Firmus’ proprietary AI Factory architecture, designed for maximum compute per watt, and developed entirely for AI-native scale. This innovation leads to maximum watts to flop and lowest cost profile for ultimate customer benefit.

“Firmus is a pure AI Factory builder,” said Dr Dan Kearney, CTO of Firmus. “We’re based in Singapore, and everything we do is focused on optimising FLOPs per watt—delivering the most compute from the least energy. That's not just a sustainability win, it's another demonstration of our capability to operate the most cost-performant compute infrastructure. Our holistic approach to AI Factory development and the underpinning architecture is tuned to maximizing computation value for our customers. Developed right here in Singapore, it’s the future of AI Infrastructure.”

This latest recognition follows Firmus’ win at the 2024 DCD Asia Pacific Awards, where the company took home the prize for Best Data Centre Infrastructure Deployment. The combination of industry-leading performance benchmarks and global customer adoption shows that the AI Factory model delivers measurable value—at speed and scale.

Read full analysis here: The GPU Cloud ClusterMAX™ Rating System | How to Rent GPUs

About Firmus

Firmus Technologies creates the most efficient AI infrastructure: AI Factories designed to operate at the max-q of AI token generation and profitability. With advanced liquid-everywhere data center technology at its core, Firmus innovates across all layers of the AI Factory, from energy grid management and thermal innovations to GPU and networking telemetry and control. This holistic approach creates AI Factories that operate at peak efficiency, delivering high uptime, throughput, and profitability with future-proofed designs.

At the heart of this infrastructure is AI FactoryOS™, Firmus’ pioneering operating system that integrates data intelligence and control across the physical AI Factory.

SMC, Firmus' AI-native cloud platform, leverages SHARP-enabled networking and NCP reference architecture, benefiting from the efficiency of our AI Factory infrastructure. SMC empowers customers to train faster, deploy smarter, and scale globally, while maximizing the value of every token generated.