Transforming supply chain performance to meet AI’s data centre demand
As technology rapidly evolves, artificial intelligence (AI) has emerged as a huge growth market, with a wide range of applications and even more unexplored potential. Our Data centre cost index 2023 shows that 88 percent of respondents reported a demand surge for data centre capacity due to the rapid increase in AI and machine learning projects and their corresponding need for exponentially more computing capacity. For data centre operators, a focus on improving the construction supply chain may be the competitive advantage they need to meet this unprecedented pressure.
By Lily Raddats, Consultant, Tammy Han, Senior Consultant and Edmund Bowe, Associate Director
In recent years, AI has become less the stuff of science fiction and more of an everyday reality, with an increasing array of real-world business applications, ranging from customer service and relationship management, to cybersecurity, accounting and inventory management. With seemingly limitless potential, this surging interest signifies a pivotal moment in the tech industry, as AI becomes more accessible to meet both business and individual needs.
The one real limitation is the availability of data centres to meet AI’s cloud computing requirements.
The need for data centres has never been greater and data centre operators must strive to build faster to remain competitive. Until recently, the cost efficiency of building and operating data centres was improving steadily. Prefabricated and modular construction techniques, technological improvements and economies of scale all played a part in this.
However, the COVID-19 pandemic increased material lead times, escalated energy prices and squeezed historically healthy margins from labour costs. This reversal corresponds with the recent increase in interest in and demand for AI. Now, in order for data centres to have an edge on the competition, they must be able to rapidly scale up and find ways to avoid passing those costs onto customers.
Within data centre construction, every penny is being tracked and coded to understand how the next one can be built more efficiently - ultimately, the cost to build hits the bottom line. As a hugely profit-driven industry, the more cost efficiently data centres can be built, the better.
With so much beyond the control of data centre operators, we have seen one area of relatively unexplored opportunity for making significant strides in improving operations: construction supply chain optimisation.
Focus on functionality and strength of the supply chain
In the world of data centres, where every decision is scrutinised for return on investment, cost considerations often dominate the landscape. However, focusing solely on cost can lead to missed opportunities and long-term risks, and it is essential for data centre operators not to underestimate the crucial role supply chains play in their success.
During the construction of data centres, their operators should shift perspective and focus on functionality and strength of the supply chain to prevent unnecessary costs and mitigate potential risks to schedule.
Data centre construction is a unique endeavor, constantly challenged by rapid technological advancements in equipment and IT infrastructure.
Innovations such as new cooling technologies, lithium-ion batteries and alternative energy sources, such as solar and 5G network adaptation, can all disrupt project delivery.
We have identified three key ways that data centre operators can transform supply chain performance to better accommodate these changes.
1. Utilise supplier performance data to inform procurement decisions and diversify supply base
It may be tempting to work with the same suppliers repeatedly, but convenience should never be confused with efficiency. Familiarity doesn’t necessarily equate to strong performance. Data centre operators would be wise to consider implementing a pay-for-performance approach to drive the highest quality standards and reflect the dynamism of the industry.
Setting KPIs to consistently measure suppliers against the four fundamental pillars of project delivery - cost, schedule, quality and safety - will provide data that can be used to benchmark suppliers and identify trends over time.
Accurate, current and trended performance data can alert you to recurring issues (for example, where schedules slip or costs increase compared to the plan), and bring to light the pain points, weaknesses and gaps in your supply chain.
With this data in hand, you can undertake a sourcing effort to expand your network in a way that proactively mitigates risk.
With fierce competition across the sector for the same services, components and equipment, diversifying your supplier base also helps avoid a single point of failure. If your primary equipment supplier experiences production issues, having a secondary supplier already identified and integrated into your supply chain should allow you to pivot without major disruptions.
2. Optimise resource allocation with capacity planning and demand forecasting
The push for data centre market growth means that both labour and product shortages are a high-risk probability, making real-time insights into local market intelligence essential. Advanced supply chain data analytics provide these insights, and demand forecasting can support efficient capacity planning with your vendors.
For example, if you proactively analyse the demand plan and match it against forecasted labour supply in your areas of construction then any gaps identified can be mitigated early to reduce the supply risks from labor shortage.
The same process can be applied for equipment or materials. For specific equipment and components, forecasting can also help you negotiate volume discounts or bulk pricing if you're confident that your demand aligns with the vendor's capacity.
3. Plan for the unknown with scenario modelling simulations
The dynamic nature of the data centre industry demands proactive planning. Scenario modelling simulations allow data centre operators to assess the impact of alternative sourcing strategies, evaluate potential schedule and cost deviations and make informed decisions.
These predictive methodologies can empower owners to negotiate with a comprehensive understanding of the potential cost implications of various contract terms.
When new opportunities arise, such as the opportunity to acquire land, simulations can help forecast the availability and timelines for long-lead equipment, enabling you to reroute materials as needed.
As the demand for data centre continues to surge in response to AI and machine learning, data centre operators must find ways to stay ahead of the competition. One of the most impactful strategies is supply chain optimisation. For data centre operators, a focus on improving the construction supply chain may be the competitive advantage they need to meet this unprecedented pressure.
