More

    Sustainable Metal Cloud Unveils Game-Changing Cooling Technology Amid AI Surge

    As the artificial intelligence revolution intensifies the demand for advanced processors and the substantial energy required to cool data centers, Sustainable Metal Cloud is stepping into the spotlight with a promising solution. The company claims its innovative technology can cut data center energy consumption by up to 50%, offering a significant edge as AI-driven demands strain the power grid.

    Sustainable Metal Cloud, known for its “sustainable AI factories,” operates its HyperCubes in Singapore and Australia. These HyperCubes feature servers outfitted with Nvidia processors, immersed in polyalphaolefin—a synthetic oil that outperforms air in heat dissipation. According to the company, this advanced cooling method slashes energy use compared to conventional air-cooling systems.

    Tim Rosenfield, co-founder and co-CEO of Sustainable Metal Cloud, highlighted the advantages of their system. “Our technology supports high-density GPU hosting essential for platforms like Nvidia’s Grace Blackwell,” he said, referring to Nvidia’s latest AI graphics processors unveiled in March.

    The Singapore-based firm also boasts that its immersion cooling technology is 28% more cost-effective to install than other liquid-based cooling solutions. HyperCubes are adaptable, designed to fit into any data center and utilize unused spaces within existing facilities. “Our containerized solution allows us to quickly adapt and open new availability zones in response to customer demand,” Rosenfield explained.

    Sustainable Metal Cloud’s expansion includes new markets such as Thailand and India. The company already collaborates with major enterprise partners like Nvidia and Deloitte. SMC is a preferred cloud partner for Nvidia, offering GPU clusters designed by the chipmaker. In July, the company announced a partnership with Deloitte, providing access to Nvidia’s GPU computing infrastructure for AI application development.

    As governments and businesses scramble to harness AI’s transformative potential, data center demand has surged. Singapore, where SMC is headquartered, is pursuing “green” data centers to manage the substantial energy consumption associated with its AI initiatives, investing over 500 million Singapore dollars (approximately $379.7 million).

    Sustainable Metal Cloud has secured funding from ST Telemedia Global Data Centres, a Temasek-backed state investor and one of Asia’s largest data center operators. The company is currently raising $400 million in equity and $550 million in debt to fund its expansion beyond Singapore, according to Bloomberg.

    The trend towards liquid cooling is gaining momentum as technology companies address the overheating risks associated with increasing performance demands. Traditional air-cooling systems, which involve large aisles for air circulation, are being challenged by more efficient liquid cooling solutions.

    Giordano Albertazzi, CEO of digital infrastructure provider Vertiv, noted that liquid cooling adoption is expected to accelerate in 2024. Vertiv’s offerings include hybrid air- and liquid-cooling systems and fully liquid-cooled data centers. At Computex Taipei in June, Nvidia CEO Jensen Huang praised Supermicro’s direct liquid-cooling technology for its energy savings and cost reductions in AI data centers. Supermicro CEO Charles Liang also emphasized that liquid cooling enhances power efficiency, performance, and reduces pollution and energy costs.

    Despite the enthusiasm, challenges remain in the widespread deployment of liquid cooling. Rosenfield acknowledged that many data centers are not yet equipped for liquid cooling. “The industry is still exploring the best ways to implement this technology, and there will likely be multiple approaches,” he said. Vertiv’s Albertazzi agreed, noting that air cooling will continue to play a significant role even as high-density AI data centers evolve.

    Related topics:

    What Is Ensemble Learning

    What Is Artificial Learning?

    What Is Deep Learning in Image Processing?

    Recent Articles

    TAGS

    Related Stories