The future of high-performance compute: How Northern Data Group is powering the next generation of AI

“The future will be defined by Generative AI. With our sustainable, state-of-the-art high-performance computing solutions, Northern Data Group is pushing the boundaries of what is possible”

Northern Data Group’s COO, Rosanne Kincaid-Smith, is a driving force behind the company’s ascent as a premium provider of High-Performance Computing (HPC) solutions – the foundation for Generative AI technology.

High-performance compute is the foundation for the technologies that will deliver accelerated results in scientific, industrial and societal advancements. Northern Data Group is leading the way in building and developing those computing solutions with sustainability in mind, and democratising access to sustainable AI technology through its AI Accelerator.

In our latest Q&A, Rosanne discusses how Northern Data Group is powering the next generation of innovation bravery through its sustainable, state-of-the-art, HPC solutions. Leveraging her experience in both international and emerging markets, Rosanne offers her perspective on the challenges from sustainability to cybercrime currently faced by the industry, and what the future may look like

1. How has the high-performance compute market been impacted by the boom in Generative AI innovation and adoption?

AI and ML technologies require a huge amount of computing power, and the exponential pace of their adoption and advancement is beyond what our traditional cloud infrastructure can support.

HPC solutions – like the ones that Northern Data provides – use clusters of extremely powerful computers working in parallel to process exponentially larger datasets and computational tasks at previously unheard-of speeds, unlocking AI and ML innovations that can revolutionise business and society overall. HPC can power far more complex tasks than cloud computing, including for example self-driving vehicle systems and drug discovery through AI-generated development of new molecules.

As a result of this demand, the HPC industry has seen demand for our services soar as companies around the world strive to meet the moment and take advantage of the vast benefits AI and ML technologies promise.

2. How do you anticipate HPC technology will evolve in the next ten years and what is Northern Data Group’s vision for delivering it in the right way?

The growth in AI and ML technologies has already proven to be transformative, even in their early stages.

This has rightly gotten a lot of attention from businesses and the press alike, but parallel transformation taking place in HPC is proving to be just as exciting.

Take the recently unveiled NVIDIA® Blackwell GPUs as an example. These can improve performance by 7-30 times, and Northern Data Group has been named as one of the first sovereign AI cloud providers of Blackwell-based cloud services and infrastructure. We are confident this will change everything. The innovation offered by this generation of products is going to help our customers immensely and is set to power the next generation of GenAI and accelerated computing.

However, we also need to support this growth sustainably. Our data centres are uniquely suited to support the high-density, ultra performant configurations of the Grace Blackwell Superchip thanks to our state-of-the-art configurations and use of liquid cooling and other cutting-edge tactics to provide the ideal data centre environment. As we look to the future, we will continue to make sizeable investments in our data centre footprint, software advancements and technology innovation to ensure we remain at the forefront of any technological development.

3. In recent months, the media has turned to focus on AI’s impact on the environment. What are your thoughts on this and how is Northern Data Group ensuring it is operating in the most sustainable way possible?

GenAI requires huge amounts of energy needed to build, train and deploy models, which in turn is driving increased power demands on data centres and impacting energy consumption globally.

This increased energy consumption can come with its own environmental costs if not managed properly, and industry needs to ensure the pace of innovation can be maintained without it becoming an obstacle to environmental goals.  We need to use these rising energy demands as a catalyst to incorporate renewable energy and invest in sustainable solutions at a faster rate.

At Northern Data Group, our data centres are powered by clean energy wherever possible, located in places that can provide clean energy or operate predominantly off renewable sources, such as our site in Boden, Sweden. This is possible because Scandinavia has excellent green infrastructure and excess energy capacity, making its Arctic environment an ideal location to build data centres, with its temperature providing efficient cooling.

Our data centres also incorporate direct-to-chip liquid cooling, rather than using the air as a coolant, which ensures the chips inside them can be kept cool and can reduce the total power consumption of a data centre by more than 10%.

4. Drawing on the topic of your panel at CogX, ‘How AI can help and hinder the problem of cyber crime’, which has increasingly become a topic of interest in recent months – how do you think advances in AI can help combat this type of threat?

AI is increasingly used by cybercriminals to escalate sophisticated and targeted attacks. However, the real value lies in the technology’s ability to mitigate the problem.

Cybersecurity professionals have been using AI for years to recognise patterns, as well as flag potential vulnerabilities and threats. GenAI unlocks a transformative shift, leveraging machine learning models that enable businesses to simulate cyberattacks and defensive strategies.

AI will be the future of cybersecurity; it can create sophisticated defences and test them against simulated attacks. The ability of GenAI to evolve rapidly and adapt to new threats as they emerge will allow for more sophisticated defence measures.  Standards around data privacy are of the utmost importance in the cybersecurity space and must be given special consideration.

The future of cybersecurity will not just be about adopting AI, but about leveraging it to augment human expertise and achieve greater security outcomes.

5. What is Northern Data Group’s goal in running its AI Accelerator? How does it play into your longer-term strategy and wider vision for the industry?

Our AI Accelerator is aimed at the world’s most innovative AI and ML focused startups. We provide those looking to supercharge their ideas and scale-up their proof of concepts with access to our AI hardware, including our NVIDIA H100 Tensor Core GPUs.

We’ve received applications from startups across a range of sectors, including health tech, ESG and cybersecurity, who are all pushing the boundaries of what is possible with AI. While applications are closed for this year, we plan on repeating the Accelerator at a later date.

Our Accelerator’s goal feeds into Northern Data Group’s wider ambition to democratise access to AI technology. The program helps advance our core value of Innovation Bravery -- providing the infrastructure and foundations necessary for companies to drive societal progress. We will enable businesses at all stages of their AI lifecycles to drive innovation in unique and transformational ways.

6. What examples of AI innovation are you most excited by that your computing infrastructure has helped enable?

Northern Data Group is differentiated by the company’s ability to offer our products as a bespoke service. We carefully select our locations in Europe which ensures in-region compliance and data sovereignty, causing us to rapidly grow our customer base. Our focus on sustainable solutions means that our customers do not have to worry about meeting ESG requirements.

We are able to adapt to the needs of our customers, working with large and small companies alike. For example, Poolside.AI builds LLMs that work to revolutionise coding. We also serve international software providers which service the likes of Cisco and IBM. On the other hand, we cater to smaller firms such as Xayn, which is building a legal copilot. One notable customer, NUA Software, created an AI-powered food recognition technology that was able to achieve a 99% accurate rate in food recognition through the use of our GPUs for training and inferencing.

GenAI has the ability to transform the work of any employee, in any sized business, and any sector. This is reflected by our diverse range of clients, whom we work with as a true technology partner.