The startup provides energy-efficient GPU cloud infrastructure designed for ultra-low latency AI training, fine-tuning, and inference. By delivering deep learning and omniverse GPU clusters at scale and lower costs, it enables organizations to enhance their AI capabilities while minimizing operational expenses.
Funding
$550M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Product
Problem
The increasing demand for AI and machine learning workloads requires significant computational power, leading to high energy consumption and escalating operational costs for organizations. Traditional GPU cloud infrastructure often lacks the energy efficiency needed to sustainably support these intensive workloads.
Solution
Sustainable Metal Cloud (SMC) provides energy-efficient GPU cloud infrastructure optimized for AI training, fine-tuning, and inference, enabling organizations to enhance their AI capabilities while minimizing operational expenses and environmental impact. SMC's infrastructure leverages advanced liquid cooling technology and is co-designed with NVIDIA to deliver high-performance computing with significantly reduced energy consumption compared to legacy cloud solutions. By offering dedicated, scalable NVIDIA GPU clusters, SMC allows customers to access the necessary computational resources without compromising on sustainability or cost-effectiveness. SMC's commitment to transparency includes releasing MLPerf-certified training power consumption benchmarks, providing customers with verifiable data on energy efficiency.
Target Audience
SMC's primary customers include AI startups, scale-ups, enterprises, government organizations, and cloud service providers (CSPs) that require cost-effective, high-performance, and sustainable GPU cloud infrastructure for AI and machine learning workloads.
Features
- Dedicated, scalable NVIDIA GPU clusters featuring H100, A100, and L40S GPUs
- Advanced liquid cooling technology that reduces energy consumption by up to 48% compared to traditional air-cooled data centers
- High-performance networking with RDMA InfiniBand for low-latency communication between GPUs
- Bare metal and managed metal options to suit different project requirements
- Integration with NVIDIA AI Enterprise platform, providing access to over 100 AI frameworks and pre-trained models
- Flexible storage solutions, including fast NVMe file storage and S3-compatible object storage
- Secure data centers with multiple Availability Zones (AZs) for high availability
- Support for burst, reserved, and on-demand GPU service models