Complete Guide to Data Center Cooling Optimization Software

What is the best data center cooling optimization software? EkkoSense Technical Account Director James Kirkwood discusses the key questions to help identify your data center cooling optimization must-have criteria.

Summary

A decision-focused guide that defines evaluation criteria, must-have capabilities, and proof points for organizations searching for data center cooling optimization software that will help them to maintain rack-level thermal compliance and support the ongoing reduction of Power Usage Efficiency (PUE) scores.

Bridging the gap between IT and M&E functions

With data centers evolving faster than ever before, operations teams are continually challenged by the need to:

• Support high-density AI workloads
• Run GPU-as-a-Service propositions
• Operate hybrid air and liquid cooling infrastructure
• Keep on top of power constraints
• And deliver against corporate Net Zero and sustainability goals

Given this, there’s never been a greater time for operations teams to understand exactly what’s going on across their data center enterprise estate in real-time. However, despite huge investments in infrastructure, data center teams increasingly find they simply don’t have access to the kind of cooling, power and capacity information and insights they need to make informed decisions about key criteria such as data center energy efficiency, HVAC optimization, thermal management, cooling monitoring and control, and PUE reduction. Despite unprecedented growth in data center investment, the reality for many organizations is that their current data center management systems simply can’t keep up with these requirements. Many legacy equipment and monitoring capabilities invariably fall short of high GPU and liquid cooling needs. For example, existing BMS and DCIM platforms cannot forecast power and cooling bottlenecks effectively to find stranded data center capacity. While legacy BMS, EPMS and other systems may capture volumes of operational data, it’s often hard to access quickly and understand and visualise what the data is saying – and more so across larger estates where there’s the added complexity of multiple incompatible M&E systems running in parallel.

Evolving Data Center Cooling Optimization software requirements

Today’s data centers can raise increasingly complex questions. Unlike traditional data center workloads, high-density AI environments are potentially much more volatile. Power demand shifts quickly, thermal conditions change in real-time and small issues can escalate fast if they’re not seen and resolved early. For organizations running or planning AI at scale, having clear real-time operational insight is now essential.

We hear repeatedly from global data center operators who, despite spending millions on traditional DCIM, BMS, EPMS and EMS systems, still don’t get the real-time operational insights they need to effectively manage their sites on a daily basis.  These traditional systems are of course excellent at providing core infrastructure control and alerting, but this is entirely different to having unique data-driven insights about key issues such as data center energy efficiency and cooling monitoring & control.

Lack of real-time visibility is simply too big a risk

Running AI workloads is clearly complex, with the need for much tighter operational tolerances and much faster responses. Previously with air cooling, for example, you had time to investigate, you could deploy temporary cooling. But with high density deployments and liquid cooling, a data center’s  tolerance window gets tighter and tighter. This places an increased burden on operations teams – with an associated requirement for more effective managing and monitoring of operations that require much quicker intervention.

Another reality teams face is that their high-density AI systems are changing every week or month. AI acceleration is shortening infrastructure planning and operating cycles, forcing operators from legacy static designs to dynamic, data-driven systems that adapt to high-density workloads. Couple in the fact that AI compute racks can easily be worth potentially millions per rack, and there’s a pressing requirement for early warning systems that can protect high value assets against critical incidents.

The stakes here are changing by orders of magnitude. When you’re investing hundreds of millions in AI data centers, lack of visibility is simply too big a risk for most operators. Relying on traditional BMS or DCIM systems alone is no longer enough, as they’re only likely to report a problem when it’s actually happening. Similarly, relying solely on engineers on-site to respond is not going to be fast enough anymore. Meeting this kind of risk profile requires new levels of granularity, responsiveness and innovation.

Optimizing data center performance for the AI era

For data center operators still held back because of the complexity of their traditional DCIM or Building Management System tools, maintaining engineers and tech specialists on site is unlikely to be fast enough for current and future high-density AI deployments. However, supporting this generation of legacy tools with a complementary AI-powered data center optimization framework can bring a powerful blend of real-time insight, predictive monitoring and absolute operational visibility.

Done right, AI-powered data center optimization can not only add value to legacy systems, but also keep on evolving through enhanced algorithms, richer estate-wide visibility, and even more intelligent advisory capabilities.

So whether it’s using AI to support hybrid liquid and air cooling in tech rooms with high rack loads, deploying AI to detect any abnormal changes in performance to provide teams with an early warning of impending issues, connecting the white space and the grey space through CDU integration, or optimizing air-side and water-side infrastructure with tailored AI chiller recommendations, there’s really no barrier now to equipping data center teams with the kind of control that DCIM solutions originally promised.

More holistic Data Center Cooling Optimization –  key evaluation criteria

At EkkoSense we believe it’s difficult to unlock the kind of performance improvements that are needed to handle greater workloads and secure energy savings unless you know exactly what’s happening in your data center in real-time. We also believe that this won’t be achievable unless data center leadership commits to bridging the gap between their IT and M&E functions.

Indeed, we see countless examples in large operators where traditional systems have proved worthless in identifying seemingly small changes in M&E plant behaviour. It’s these kind of issues that need to be picked up and resolved before they turn into site impacting issues. Not surprisingly, operations teams need as much early warning as possible so they can take corrective action by moving workloads and avoiding costly downtime.

While the latest digital transformation and AI workload applications may run on leading edge platforms, it’s still the traditional facilities management teams that manage and maintain the building and the critical supporting infrastructure within it. However, most IT teams have little interest in the underlying M&E infrastructure that provides the power and cooling that enables their services to run. Because of this, it’s not unusual to see expensive power and cooling resources being used inefficiently. Excess energy usage not only gets in the way of corporate net zero initiatives, but also potentially places organizations at risk when critical resources suddenly become depleted or unavailable.

EkkoSense’s Top 12 recommended key Data Center Optimization needs checklist

That’s why any effective Data Center Optimization (DCO) approach needs to satisfy a number of core evaluation criteria. Any successful DCO solution should support and address the following 12 key evaluation criteria requirements at a minimum:

1. Delivering Complete Operational Visibility in Real-Time and historically
2. Identifying the changes needed to Reduce Cooling Energy Costs
3. Matching cooling with IT workloads to Release Stranded Cooling Capacity
4. Using advanced analytics to Remove Thermal and Power Risks
5. Unlocking Immediate Carbon Savings to support Net Zero programs
6. Reducing administrative overload by Automating Full ESG Reporting
7. Simplifying data center optimization using an Intuitive 3D Digital Twin
8. Ability to Ingest Data from Multiple Sources for greater granularity/accuracy
9. Embedding the latest AI & Machine Learning to disrupt traditional DCO models
10. Pick up seemingly Small Changes in M&E Plant Behavior – identifying problems before they turn into site-impacting issues
11. Leverage the latest SaaS and AI capabilities to Achieve an ROI < 12 months
12. Bridging the Intelligence gap between Core IT and M&E facility functions

12 Key Question to help identify your Data Center Cooling Optimization Must-Have Criteria

Whether you’re a CEO, CFO, CIO, COO, Data Center Manager, FM Specialist, Sustainability Leader, or part of the operations team, you’re always going to need answers to your critical data center and IT infrastructure questions. 12 key questions that will help you to identify and prioritize your Must Have Criteria include:

1. How can I access the tools and insight needed to remove power and thermal risks across my data centers?
2. How can I find out if I have any stranded power or cooling capacity that I can use before requiring further investment?
3. How easy is it to produce a clear single pane view of my extended data center estate?
4. How do I focus in on where I could cut down on my data center cooling?
5. How do I find out about potential downtime issues without having to wait for an SLA breach?
6. How can I be sure that my multiple co-location partners are delivering against their stated SLAs?
7. If you’re a co-location service provider, how do you know how much power each of your customers is using in each of their racks?
8. It’s taking me too much time to carry out all my ESG reporting – how can I make this process more efficient?
9. How can I sensibly prioritize all my data center equipment and infrastructure maintenance?
10. How can I ensure uptime while reducing maintenance costs
11. How can I be sure that my high value assets such as high density AI compute engines are fully protected?
12. How can I work out my data canter’s carbon footprint reduction and contribution to corporate net zero and sustainability targets?

Building your Next Generation Data Center Cooling Optimization Platform

As data centers become more complex, Data Center Optimization tools that can illuminate complex power, cooling and capacity relationships, install quickly, and deliver the right insights are in demand. Not surprisingly, there’s a premium on those optimization platforms that can extend operational intelligence across both the data center white space and the broader mechanical grey space.

Use our EkkoSense Data Center Cooling Optimization Platform matrix to help zero in on key features that will help inform your data center optimization software selection.

Bringing it all together with EkkoSense

Too  many organizations still find that their legacy data center cooling optimization systems don’t provide them with the critical cooling, power and capacity data they need to make the right decisions. EkkoSense resolves this by making your invisible data visible.

Our innovative and quick-to-value EkkoSoft Critical AI-powered 3D visualization and analytics optimization software tells you exactly what’s going on in real-time  –  and exactly what you need to do next to optimize performance further. So whether you’re an Edge or Distributed IT operation, a Legacy site, a Colocation provider, or an AI Factory or Hyperscale facility, our software gets you straight to the data you need to run your sites more effectively.

Over a decade of AI-based data center cooling optimization experience

With the AI scramble in full swing many organisations, not surprisingly, are now racing to bolt AI features onto their legacy DCIM platforms. However these kind of features are proving difficult to realize without AI already being embedded into your technology stack. At EkkoSense we’re able to address this having built up over a decade of AI-based data center optimization experience, drawing on Level 1 AI expertise and turning over 50 billion AI data points into clear, auditable operational optimization decisions.

Capabilities such as the industry’s first embedded AI cooling advisory tool, or a machine learning powered model that reveals exactly which cooling units serve which racks, didn’t just appear as bolt-on features – they evolved over more than a decade of focused R&D. These and the next generation of AI-powered data center optimization functionality will prove critical as data center operations teams move from simply responding to incidents to enabling the real-time management of the whole facility’s health.

Unlocking value with EkkoSoft Critical

Our award-winning, light-touch AI-enabled data center optimization platform provides an intelligent, AI-powered wrapper across all your key IT, data center and M&E systems. Focused across Cooling Energy Spend Optimization, Capacity Management, Outage Prevention, Compliance Reporting, Ease-of-Use, AI/Machine Learning Capabilities, and Enhanced Visibility, our SaaS solution complements your existing M&E infrastructure and giving you the control you need to introduce new levels of data center optimization. This allows you to:

• Remove thermal and power risks  – with improved cooling compliance and reduced risk through continuous monitoring and AI-enabled insights such as applying powerful anomaly detection for early warnings before critical equipment failures occur
  
• Reduce cooling energy costs – with unique real-time cooling advisor analytics that support complex hybrid liquid and air cooling deployments – leading to up to 30% data center cooling energy savings
  
• Release critical power and cooling capacity – with room/site/estate level power monitoring and capacity planning tools and real-time operational insights
  
• Reduce your carbon footprint – with site level cooling optimisation algorithms and continuous embedded hybrid cooling advice to align with Net Zero goals
  
• Solve ESG Reporting headaches – with simple, straightforward reporting that automates ESG Reporting

These aren’t theoretical benefits –  they are outcomes documented by major global data center operators who rely on EkkoSense every day.

Find out how EkkoSoft Critical software is easy-to-use, deploys in days, and provides an unrivalled 12-24 month payback, with an instant free demo.