Enquire nowLog in

Singapore Cloud Region

H100 GPU clusters have arrived in Singapore. The world's most energy-efficient AI infrastructure stack delivers large-scale H100 clusters, with twice the energy efficiency and 80% better TCO than legacy clouds.

  • Highlights
  • Tech Specs
  • Use cases
  • Sustainability
  • Lower TCO
  • Get in touch
Highlights

1,024+ H100 GPUs per cluster. RDMA InfiniBand. Hosted in Singapore,
globally available

SMC's Singapore Region delivers flexible, cost-effective AI tools — thanks to the world's most energy-efficient AI infrastructure stack.

Singapore Region hosts two geographically diverse Availability Zones (AZs) for high-availability options, connected across the island through SMC's multi-terabit Secure Access Network.

Available instance types include H100 & L40S clusters, with zero-compromise performance that scale to over 2,048 GPUs per cluster.

Tech specs

A hyperscale cluster of SMC GPU infrastructure

Clusters can be rented for enterprise, government and CSPs in small units or as a whole cluster. Use SMC as bare metal or deploy tools from partners like NVIDIA to get started with your AI project today.

Instance: H100
Network Fabric: Built for performance
Cluster Scale

NVIDIA certified H100 HGX instances with NVSwitch are networked into powerful clusters for large-scale AI development in Singapore.

Delivered via managed metal or with an AI development stack delivered through NVIDIA AI Enterprise - with a full suite of AI tools, frameworks and pre-trained models.

Enquire now
GPU

H100 x 8 80GB SXM

Link

NVSwitch 900Gbps

Network

ConnectX-7 Bluefield 3 DPU

CPU

Intel Xeon Platinum 8462Y+ 128 vCPU

Memory

2,048GB DDR5 4800MT/s RDIMMs

Storage

30TB NVMe

Secure and powerful GPU clusters in Singapore

H100 clusters have arrived in Singapore. Up to 70% more cost-effective, powered by 100% immersion cooling.

Built to power Singapore's next-gen AI

Our SIN01 & SIN02 AZs offer the tools and capacity to fuel future AI projects of all scales.

NVIDIA frameworks

Building for AI takes a quality platform, and quality tools. As an NVIDIA CSP Partner, SMC instances can be delivered with NVIDIA AI Enterprise platform included, featuring 100+ frameworks, pre-trained tools and enterprise support and security directly from NVIDIA.

Complex AI Workloads

All SMC clusters feature cutting edge NVIDIA GPUs and network fabrics, benchmarked to perform on some of the world's largest and most complex AI workloads. Perfect for rapidly testing, training and refining AI models across all industries.

Bare Metal scale

For very-large enterprise, government and CSPs, use the raw power of SMC's bare metal configuration. Take control of the entire cluster to secure rare access to the region's most available, sustainable and cost effective GPU cluster - entirely for you.

Build production AI faster on Singapore's most performant AI computing cluster.

Advanced Liquid Cooling

Unmatched energy efficiency

Years of R&D have led to some of the most sustainable cloud technology in the world, emitting up to 48% less carbon than other cloud providers, whilst hosting more powerful workloads to democratize access to AI computing power.

LESS CO2
BETTER TCO
Lower TCO

The sustainability discount. Energy efficient technology improves by over 80% versus legacy GPU infrastructure.

Both Singapore AZs feature SMC's advanced cooling technology. Even in hot and humid Singapore, SMC's H100 clusters run within a 1.02 pPUE environment, using materially less energy than legacy clouds hosted locally.

Passing this saving onto you makes SMC's H100 over 80% more cost effective versus other general clouds in Singapore.

Get in touch

Make the transition to SMC seamless. We’re here to assist you. Fill out the form below and we’ll be in touch soon.

Thank you for your enquiry.

The SMC team will be in touch shortly.