Amazon Web Services Unveils Custom Cooling Solution for Next-Gen Nvidia GPUs in AI Infrastructure

2 Sources

Share

Amazon Web Services has developed a custom cooling solution, the In-Row Heat Exchanger (IRHX), to manage heat from dense GPU configurations in AI workloads, particularly for Nvidia's latest chips. This innovation allows AWS to offer high-performance AI infrastructure through its new P6e instances.

AWS Innovates to Cool Next-Gen AI Hardware

Amazon Web Services (AWS) has unveiled a groundbreaking solution to address the cooling challenges posed by the latest generation of NVIDIA GPUs used in artificial intelligence workloads. The company's custom-designed In-Row Heat Exchanger (IRHX) represents a significant advancement in managing the thermal output of high-density GPU configurations

1

.

The Cooling Conundrum

Source: Benzinga

Source: Benzinga

As the AI boom accelerates, the demand for more powerful GPUs has led to increased energy consumption and heat generation. NVIDIA's latest GPUs, particularly the GB200 NVL72 system which packs 72 Blackwell GPUs into a single rack, require innovative cooling solutions beyond traditional air cooling methods

1

.

Dave Brown, VP of Compute and Machine Learning Services at AWS, explained that commercially available cooling equipment was insufficient for their needs:

"They would take up too much data center floor space or increase water usage substantially. And while some of these solutions could work for lower volumes at other providers, they simply wouldn't be enough liquid-cooling capacity to support our scale."

1

AWS's Innovative Solution

Rather than constructing entirely new liquid-cooled data centers, which would have been time-consuming and costly, AWS engineers developed the IRHX. This solution can be integrated into both existing and new data centers, providing the necessary cooling capacity without requiring a complete infrastructure overhaul

2

.

P6e Instances: Powering Next-Gen AI

The new cooling technology enables AWS to offer P6e instances, which provide customers with the computing power needed to train and deploy large-scale AI models. These instances are based on NVIDIA's GB200 NVL72 architecture, which allows the 72 GPUs to function as a unified compute system

2

.

Industry Context and Competition

AWS's innovation comes amid fierce competition in the AI infrastructure space. Microsoft and CoreWeave have previously offered computing clusters based on NVIDIA's GB200 NVL72 architecture. Meanwhile, Intel has positioned its SuperFluid cooling technology as a competitor, claiming it can handle thermal loads up to 1,500 watts

2

.

Implications for AWS and the AI Industry

Source: CNBC

Source: CNBC

This development reinforces AWS's position as the world's largest cloud infrastructure provider. By creating custom hardware solutions, AWS can potentially reduce its dependence on third-party suppliers and improve its bottom line. In the first quarter, AWS reported its widest operating margin since at least 2014, contributing significantly to Amazon's net income

1

.

As the AI industry continues to evolve rapidly, innovations in cooling technology will play a crucial role in enabling the next generation of AI applications and research. AWS's IRHX solution demonstrates the company's commitment to staying at the forefront of this technological revolution.

TheOutpost.ai

Your Daily Dose of Curated AI News

Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Spend less time searching for the latest in AI and get straight to action.

© 2025 Triveous Technologies Private Limited
Instagram logo
LinkedIn logo