As rack power densities continue to increase across hyperscale, neocloud, and enterprise data centers, operators are under growing pressure to improve cooling performance while reducing energy consumption.

This trend is being accelerated by AI workloads, GPU clusters, and other high-density computing environments that generate unprecedented thermal demands.

These rising thermal demands are creating a need for new approaches that can improve both cooling performance and energy efficiency.

Recent testing of a new nanofluid for data center cooling showed promising results, with a 30-50 percent reduction in fluid approach temperatures in both the coolant distribution unit (CDU) and evaporator, along with impressive improvements in compressor efficiency.

The results indicate significant advances in both cooling and energy efficiency for hyperscale, neocloud, and enterprise data centers, holding the potential for dramatic reductions in power usage effectiveness (PUE) and millions in energy savings.

On behalf of a major hyperscale client, Schneider Electric conducted the tests on a closed-loop data center cooling system in its manufacturing hub laboratory in Conselve, Italy. The tests compared the thermal capabilities of a conventional PG 25 cooling solution to PG 25 combined with a graphene nanofluid called AQUENE, developed by Blue Snow Frontier Technologies.

What is AQUENE, and how does it improve data center cooling?

AQUENE is a nanofluid that contains suspended and functionalized graphene nanoparticles. Because graphene has a thermal conductivity that far exceeds conventional chilled water, the resulting fluid transfers heat far more efficiently from servers to heat exchanger surfaces.

The practical effect is that a data center cooling system using AQUENE can remove the same amount of heat from servers while consuming less energy than PG 25 alone. (Note PG 25 was used because temperatures in that part of Italy can dip below freezing in winter, and glycol acts as an anti-freeze. In warmer climates, glycol would not be required.)

AQUENE test parameters and results

The tests evaluated AQUENE under a range of realistic data center operating scenarios, including:

• Three technology cooling systems (TCS) supply fluid temperatures: 20°C, 25°C, and 30°C.
• Three cooling load levels in the TCS loop: 100 percent (100kW), 75 percent (75kW), and 50 percent (50kW).

This multi-variable testing design ensured performance gains were not isolated to a single operating point. The tests demonstrated a clear reduction in approach temperature at both the CDU and the evaporator (i.e., the FWS, or Facility Water System), indicating improved heat transfer across the cooling system for all three supply temperatures.

Approach temperature represents the minimum temperature difference between two fluids at the point of closest thermal contact in a heat exchanger. It is the driving force for heat transfer at that location. A smaller approach temperature means more effective heat transfer.

At the CDU, the approach temperature decreased between 0.7K and 1.7K, depending on load and set point temperature. At the evaporator, temperatures decreased between 0.3K and 1.2K. These lower approach temperatures demonstrate closer thermal coupling between the working fluids and more effective heat exchange.

In other words, these tests show that AQUENE enables the cooling system to remove the same amount of server waste heat while using less energy than PG 25 alone. The results also indicate that the solution has the potential to support higher-density IT loads without increasing energy consumption.

The reduced approach temperatures in the evaporator also resulted in significantly increased compressor efficiency across all three operating capacities evaluated. This consistent improvement indicates the efficiency gain was sustained under varying conditions, not just at a single test point.

Taken together, the reduced approach temperatures and the resulting improvement in compressor efficiency translate into more efficient overall cooling-system operation, enough to reduce PUE by up to 10 percent to 15 percent under similar operating conditions as the test environment. In environments where even a modest improvement in efficiency can mean millions in energy cost savings, that kind of reduction is significant, to say the least.

Why graphene? The science behind the performance

The performance improvements speak to the power of graphene as a nano additive in improving thermal conductivity.

Thermal conductivity is typically measured in watts per meter-kelvin (W/mK). A larger number means heat moves through the material more easily. Graphene has a thermal conductivity exceeding 5,000 W/mK, whereas water performs at about 0.55 W/mK.

This exceptional thermal conductivity derives from graphene’s structure, which is a single layer of carbon atoms arranged in a honeycomb lattice. By dispersing graphene nanoparticles within a water base using patented technology, Blue Snow created a nanofluid that capitalizes on graphene’s superior heat transfer properties. The resulting fluid transfers thermal energy more efficiently from the fluid to the heat exchanger surfaces within the chiller.

What are the implications for hyperscale, neocloud, and enterprise data centers?

The hyperscale client that requested the testing is now looking ahead to field testing AQUENE in an operating data center environment in Europe in Q3 2026. The expectation is that hyperscale operators dealing with rapidly increasing rack densities can use AQUENE to extract greater cooling efficiency from existing infrastructure without significant capital expenditure.

The results also show promise for neocloud and enterprise data centers. As data center owners try to save on water consumption, many are moving toward closed-loop cooling towers. AQUENE can be used in such closed-loop systems as well.

What are the next steps for AQUENE deployment?

As AI drives higher rack densities along with greater power and cooling demands, even incremental efficiency gains can have a meaningful impact at scale.

AQUENE has already shown excellent results in sites such as office buildings and hospital central cooling plants. The hyperscaler test results suggest AQUENE could deliver more than incremental improvements for data centers as well.

With the potential to reduce PUE by 10-15 percent, AQUENE could help hyperscalers, neoclouds and enterprise data center operators improve energy efficiency, lower energy costs by millions of dollars, and better manage the growing thermal demands of AI infrastructure.

Schneider Electric is continually pursuing and supporting innovations such as AQUENE to drive energy efficiency in its own business and for its customers, including data center owners and operators. The ownership in testing this new coolant innovation demonstrates this commitment to advancing technologies that deliver measurable efficiency gains, reduce energy consumption, and create business value.

Schneider Electric also leads by example. For the third year in a row, it was named number one out of 750 companies ranked on the TIME Magazine list of the World’s Most Sustainable Companies.

Explore Schneider Electric’s latest innovations in liquid cooling and discover how next-generation cooling technologies can help support AI-ready, energy-efficient data centers.