Urgent need for universities to retain supercomputing expertise

The great research exodus: What should universities be doing to retain their brightest supercomputer brains?

Universities are pivotal to the United Kingdom’s research industry, providing innovation to a whole host of sectors. Supercomputers and high-performance computing (HPC) have fed into this innovation for decades. However, recent worrying trends suggest an exodus of talent and the search for new horizons in other industry sectors.

HPC staff working in research and education often possess a unique set of skills and experiences that make them highly sought after in the corporate world. The transition from academia to the corporate sector is an increasingly common career move, and there are several compelling reasons behind this trend.

UK universities have a challenge on their hands; not just in retaining their very best talent, but in helping to provide clearer pathways and better opportunities for young, fresh candidates through study options and then into research and HPC.

However, according to the latest statistics from the Higher Education Statistics Agency, there has been a 45% growth in the number of individuals transitioning from academia to industry – from 1,490 to 2,160.

There are various factors that impact upon the decision to move – intersectoral mobility, as it is often referred to – ranging from seeking financial and job security on a personal level through to external impacts such as funding challenges, limited career progression and work-life balance … or imbalance, even.

Likewise, while the increased presence of academics in industry offers significant benefits by allowing companies to access cutting-edge knowledge and transferable skills, it’s crucial to try to understand if the moves are involuntary, such as a contract ending, or to seek personal gain and stability.

In addition, HPC professionals in research and education develop a specialised skill set that is in high demand. They have expertise in parallel computing, cluster management and the optimisation of complex algorithms. These skills are not only transferable but extremely valuable in certain corporate environments where big data analysis and computational modelling are becoming crucial.

Furthermore, researchers in HPC have honed their problem-solving abilities through the constant need to optimise code, manage large datasets and address complex scientific challenges. These skills are applicable in the corporate world, where professionals are often tasked with resolving intricate issues, typically across supply chain management, defence, finance, healthcare, engineering and a plethora of alternative sectors.

And let us not forget, many universities and research institutions have collaborations with corporate partners. This provides HPC staff with exposure to alternative work cultures and projects, making the transition appear to be attractive, and potentially seamless too.

What can be done to stem the bleed?

Clearly, it’s unrealistic to expect academia to match all of the financial benefits and career prospects that the corporate world can offer, and the lure will always exist. And who can blame someone for wanting to move to, say, pharmaceuticals or Formula One, which without HPC these past 20 years would look very different, and all the glamour it brings compared to the uncertainty of budget constraints and the never-ending funding merry-go-round.

But UK universities can still, with every justification, hold their ground as hotbeds of research and innovation for many years to come and perhaps therefore the government needs to focus upon the added value that academia brings and invest in ways to add security and create incentives and a better work-life balance for researchers.

Attracting the interest, and the money

Universities, and the UK government, could certainly do more to both ‘PR’ the work of researchers in academia, and to incentivise them.

In October 2023, the National Centre for Universities and Business (NCUB) released a report, Attracting International Investment in Research and Development (R&D), which outlines a strategy for attracting foreign investment into R&D and suggests actions to win more overseas investment.

The report states: “Coordination therefore between the supplies of skills, in terms of both higher education and further education, and efforts to encourage clusters, with higher education at the heart, is vital for the attraction of high-tech FDI [foreign direct investment].”

Dr Joe Marshall, the chief executive of NCUB, was quoted as saying: “Although the UK draws a significant amount of international investment, now is no time for complacency.

“If the government is serious about growing the UK economy through greater research and innovation, we need to see a fundamental step change in levels of private R&D investment into the UK from overseas. However, despite various strategies, reviews, incentives and policies, private R&D investment is broadly flatlining.” And I would agree.

While this report focuses upon the need for collaboration between business and academia to enable the UK to remain a progressive competitor in the global market for R&D investment, the role of UK universities is vital for this, and for the benefits to be felt across UK industry. Fundamental to that success, surely, is skilled researchers, underpinned by experience of HPC.

Furthermore, UK universities are facing growing challenges in bridging the funding gap for ‘full economic costs’ that remains uncovered by UK Research and Innovation grants and other research grant sources. This often leads to university budget holders having to cross-subsidise research from other areas, which, of course, has a knock-on effect, but this is clearly ultimately unsustainable.

Likewise, HPC funding to support research can be a minefield, depending upon who’s influencing, or commissioning, the decision to invest.

In a recent report Red Oak published, Incorporating the Cloud into the HPC Mix, we tackle this very dilemma and examine how universities should look at total cost of ownership when investing in HPC, rather than eyeing a fixed or depreciating asset, to fully support their research departmental staff.

Without HPC, we can’t access the data at the speeds and accuracy required, so this is one area that universities can use to help bolster their research departments, and reassure their staff.

The need for pathways

To my knowledge, there aren’t many universities in the UK, other than Edinburgh, providing high quality HPC courses. A trick is definitely being missed here as these courses could, and should, provide clear pathways for the next generation of HPC talent coming through their doors, and into research. If this were to be the case, the bleed would not be so severe at the top end, given the conveyor belt of fresh young talent coming through.

Similarly, more could still be done by universities to encourage students into HPC, and computer sciences in general. Open days, and STEM days, that encourage pupils considering options and sixth formers into the sector should be focused not just upon the virtues of the course, and the merits of their teaching, but on how it opens up wider horizons into the sectors we have already discussed.

It should be made clear that a career in HPC can lead to a person having a significant impact upon our future economies, health and even the future of our planet.

Finally, it’s noteworthy that there is such an imbalance in gender roles in the HPC sector. When I first started out in HPC some 30 years ago, I’d guess that the male-female split in HPC was pretty equal.

That has swayed dramatically now, and while there has been a recent upsurge in women enrolling in computer science courses, male students still outnumber females by 4.3 to one, according to 2022 figures. There is clearly still a lot of work to be done by UK universities to encourage women into HPC and computer sciences.

High-performance computing staff in research and education possess a wealth of skills and experiences that makes them highly attractive to the corporate world. Their specialised knowledge, problem-solving abilities, strong work ethic and adaptability make them valuable assets for companies across a range of industries.

As the demand for data-driven decision-making and computational expertise continues to grow, the migration of HPC professionals to the corporate sector is likely to persist and to benefit both sides through the exchange of knowledge and innovation. But action needs to be taken to encourage new talent into academia with a clear and incentivised pathway to allow UK universities, research and HPC to also thrive.

Owen Thomas is a partner at Red Oak Consulting, which helps companies across the globe with their high-performance computing (HPC) solutions. Founded to fill a knowledge gap in the IT sector, Red Oak Consulting focuses solely on HPC and cloud services.