Northern Data Cloud Services has joined the NVIDIA Partner Network as a Preferred Partner for Cloud Services in Europe. Northern Data’s sustainable data centers provide a GPU compute infrastructure based on NVIDIA A100 Tensor Core and RTX A6000 GPUs, primarily targeting artificial intelligence (AI) and professional visualization development teams. Northern Data helps these teams solve research and development assignments that require high processing power as well as data-driven business models. This infrastructure creates the platform for automated and semi-automated workloads in fast-growing segments such as image, video and data analysis, natural language processing (NLP), rendering, and machine learning (ML).
Aroosh Thillainathan, CEO of Northern Data, states: “We are delighted to be an NVIDIA Partner Network Preferred Cloud Service Provider. The market is just getting started in the development of GPU cloud computing, yet we have the capability to offer large volumes of GPU computing power, tailored to the growing needs of our customers based on our own HPC data center infrastructure. We run our climate-neutral data centers in Scandinavia with a focus on the highest levels of energy efficiency. As a European supplier, we also guarantee our customers data sovereignty and security in line with European standards to meet their compliance requirements.”
“Combining our powerful full-stack accelerated computing platform with Northern Data’s user-friendly and climate-neutral concept marks a milestone in the European cloud services market,” says Markus Hacker, Regional Director of Enterprise Business DACH at NVIDIA, and continues: “Together, we’ll drive demand for these new, cloud-based HPC offerings.”
A powerful overall package
With a large number of NVIDIA GPUs, Northern Data will provide start-ups, scale-ups, established companies and system integrators with massive computing capacity for data-intensive workloads on demand. With Northern Data, customers will benefit from a tech stack that guarantees high interconnectivity and absence of latency between individual GPUs, as well as within one tenant for maximum scalability and flexibility. In accomplishing this, Northern Data is relying on the latest networking technology from NVIDIA, all-flash storage solutions from PureStorage and an external, redundant 100 Gbit internet connection. High security standards, low complexity of the network configuration and the simple scaling of workloads thanks to the support of Cloud-init complete the offerings.
To simplify developer workflows and optimize performance on GPU instances, NVIDIA offers NGC™, a Hub of GPU-optimized AI and HPC software including enterprise-grade containers, frameworks, pre-trained models, Helm charts and industry-specific software development kits (SDKs) for data scientists, developers and DevOps teams to build and deploy solutions faster.
This end-to-end approach — from the choice of location to the construction and operation of the data center — and Northern Data’s focus on cloud-native applications and the formulation of transparent flavors guarantee Northern Data customers cost-efficient use of sought-after HPC resources with no surprise follow-on costs. This is enabled by Northern Data’s sustainable data centers with exceptionally high power usage effectiveness (PUE) in Sweden (PUE value: 1.04 in 2021) and Norway (PUE value: 1.15).
“The underlying technology and framework conditions in which it is provided are crucial for our offerings. NVIDIA GPUs speed up large, demanding workloads as they can process data volumes in the petabyte range much faster than traditional CPUs. With the exceptional performance of NVIDIA GPUs, large-scale simulations can also be processed faster than ever before,” explains Christopher Yoshida, President and CFO of Northern Data.
Cost transparency for wide-ranging application scenarios
The current cloud offerings target a wide variety of application scenarios for GPU computing. For example, they include fast, efficient image and video analysis for media content creators, rapid rendering of complex models and 3D simulations for research and development teams, real-time analysis and evaluation of large data volumes for business decision-makers or even the accelerated training of AI and ML models.