Contact us

The future of data centers demands advanced cooling

Posted on
November 1, 2024

Liquid cooling advantages over other methods:

Enhanced cooling efficiency

Enhanced cooling efficiency
Liquid cooling systems can maintain lower temperatures more effectively than air cooling. This is crucial for high-performance computing environments where even slight temperature increases can lead to performance degradation.

Space optimization

Space optimization
With the ability to cool more densely packed servers, liquid cooling allows data centers to maximize space.

Renewables and energy savings

Energy savings
Liquid cooling systems can significantly reduce energy consumption. By minimizing the need for extensive air conditioning systems, data centers can lower their overall energy costs, contributing to a more sustainable operation.

Quality Assurance

Greater reliability
By maintaining optimal temperatures, liquid cooling can extend the lifespan of hardware components. Moreover, consistent cooling can prevent thermal throttling, where processors reduce their performance to avoid overheating.

Scalability

Scalability
As AI and other technologies continue to evolve, data centers must be able to scale efficiently. Liquid cooling systems can be designed to accommodate future growth, making them a flexible solution for expanding facilities. Modular designs allow for easy integration of additional cooling units as server loads increase.

Liquid cooling solutions for modern data centers

Transitioning to liquid cooling technology requires careful planning and consideration. There are several types of liquid cooling systems available, including:

  • Direct-to-chip cooling:
    This method involves circulating coolant directly to the heat-generating components, such as CPUs and GPUs. This approach maximizes cooling efficiency and minimizes the risk of overheating. The evolution of direct liquid cooling has focused on two main technologies—microchannel and microconvective cooling.
  • Immersion cooling:
    In this system, servers are submerged in a non-conductive liquid that absorbs heat. This method is effective for high-density environments, but there can be concerns about environmental impact, leakage, and the system being a single point of failure.

Enhancing power delivery for AI servers with VPMs

As AI workloads demand increasingly powerful GPUs, Flex has developed Vertical Power Modules (VPMs) that are strategically positioned directly beneath the processor on the printed circuit board (PCB). This innovative design enhances power delivery efficiency by minimizing losses in the power delivery network, allowing for optimal performance in AI applications.

Scaling new heights with vertical power delivery

Flex’s expertise in integrating VPMs with advanced cooling solutions, particularly liquid cooling systems, addresses the thermal challenges posed by these high-power components.

Our VPMs are specifically designed for compatibility with various processors, facilitating efficient PCB trace routing and thermal management. Their low-profile design and large surface area make them ideal for direct-to-chip liquid cooling, significantly improving heat dissipation compared to traditional air cooling methods.

Flex has a long history with liquid cooling in the industrial and automotive markets. The company has formed partnerships with leading liquid cooking companies to provide advanced cooling solutions at both the board and rack level. Coupled with our global manufacturing and lifecycle services, we are well-equipped to deliver data center IT and power infrastructure solutions integrated with liquid cooling to help customers solve critical challenges with power, heat, and scale.

Leveraging innovative microconvective liquid cooling technology

Flex’s partnership with JetCool delivers advanced liquid cooling solutions specifically designed for AI servers and high-density computing.

Microchannel vs JetCool microconvective cooling architecture

This collaboration introduces a new line of liquid cooling-ready servers that comply with Open Compute Project (OCP) specifications, utilizing patented microconvective cooling® technology, which effectively targets and cools processor hot spot. This ensures optimal performance and reliability for demanding AI workloads.

The ever-increasing power demands of AI servers require reliable and innovative cooling systems. By integrating cutting-edge cooling and server technology, we can support a wide range of AI workloads and high-performance computing to help our data center customers scale.

Paving the way for tomorrow’s data centers

Liquid cooling technology is not just a trend; it is a necessary evolution in the way we manage power and heat in data centers. As we look to the future, it is clear that liquid cooling will play a pivotal role in shaping the next generation of data centers, enabling them to thrive in an increasingly data-driven world.