Zero-Emissions AI Infrastructure: A Case Study of atNorth and Neu.ro

The customer, atNorth, is a leading Nordic data center services company offering environmentally responsible, power-efficient and cost-optimized hosting facilities and high-performance computing services. They offer sustainable and extremely scalable cloud HPC resources, fully enabling customers to focus on their simulation applications and calculations without having to worry about the underlying HPC infrastructure. The company operates in the zero-emissions IT infrastructure and cloud services industry. They are committed to providing solutions that are not only efficient and effective but also environmentally friendly.

The cloud revolution is in full force, with AI being a major driver of new cloud adoption. IDC estimates that fast-growing AI workloads will account for up to 50% of the total cloud market by 2025. However, this progress comes at a significant cost to the environment. Information and Communications Technologies (ICT) already account for an estimated 9% of total global electricity use, a figure that could more than double by 2030. AI’s portion of electricity consumption is growing much higher than other technologies. Deep Learning models and the data sets they train upon are growing at an extraordinary rate. In less than 5 years, the leading language model will have increased in size by over 100,000x. The competitive challenge presented by hyperscale CSP AI development platforms (Azure Machine Learning, AWS Sagemaker, GCP AI) required an immediate response. Neu.ro provided a market proven platform with unique advantages.

To facilitate AI workload growth and satisfy atNorth’s current and future AI needs, Neu.ro proposed to install an oem cloud AI orchestration and interoperability solution to reside natively on atNorth’s zero emissions cloud service. The unique advantages of the platform are easy and fast access to the computing infrastructure from a CLI or menu system, access control and permissions for authorized team members, orchestration of both cloud and on-prem compute resources, automation of workflows and protection of AI assets and artifacts at all stages of the ML lifecycle. The platform also integrates a wide selection of best-in-breed AI/ML toolsets covering the entire ML life cycle. Following an accelerated timeline, Neu.ro successfully installed, tested and launched turnkey AI/ML services for atNorth, allowing them to leverage their existing corporate client base by offering a compelling alternative to AWS Sagemaker and offering ease and flexibility for creating, training, testing and deploying models.

• The solution provided by Neu.ro has allowed atNorth to offer a compelling alternative to other major AI development platforms such as AWS Sagemaker. This has enabled them to leverage their existing corporate client base, offering ease and flexibility for creating, training, testing and deploying models. The platform also provides additional functionality permitting multi-cloud and hybrid cloud architectures and a selection of ready and pre-integrated AI/ML products, apps and APIs for use. This provides users with advantages in terms of reduced cost, time, difficulty and risk for AI development projects. Furthermore, the platform includes an integrated toolset to manage the environmental impact of AI programs, allowing users to monitor the CO2 footprint of every job, model, project, team and optimize compute usage to balance performance and environmental impact.

• 100% powered by geothermal & hydro generation

• AI’s portion of electricity consumption is growing much higher than other technologies

• In less than 5 years, the leading language model will have increased in size by over 100,000x