This Nvidia partner will help cut AI data center energy consumption in half

The boom in artificial intelligence is driving demand for more powerful processors and the energy needed to cool data centers.

That’s the opportunity for Sustainable Metal Cloud, a data center company that runs “sustainable AI factories” made up of hypercubes in Singapore and Australia.

The Hypercube contains servers with Nvidia processors immersed in a synthetic oil called polyalphaolefin, which dissipates heat more efficiently than air. The company says the platform can reduce energy consumption by up to 50% compared with traditional air-cooling techniques commonly used in data centers.

“It enables high-density hosting of GPUs, which is what’s needed for platforms like (Nvidia’s) Grace Blackwell,” Tim Rosenfield, co-founder and co-CEO of Sustainable Metal Cloud, said of the new generation of AI graphics processors. NVIDIA It was announced in March.

The Singapore-based company also said its immersion cooling technology is 28% less expensive to install than other liquid-based solutions. Hypercube is designed to be installed in any data center and can be deployed in unused space within an existing data center.

“Because our solution is containerized, we can move it anywhere quickly, and we can open new availability zones as demand from our customers grows…” Rosenfield said.

SMC is also expanding into other markets such as Thailand and India, he said.

The company already counts Nvidia and Deloitte among its key enterprise partners. SMC is Nvidia’s preferred cloud partner for computing and AI, providing GPU clusters designed by the chip giant. In July, SMC announced a partnership with Deloitte to provide access to Nvidia’s GPU computing infrastructure to help the consultancy’s clients build AI applications.

Governments and businesses are rushing to capture the transformative impact of AI, resulting in a surge in demand for data centers.

Countries such as Singapore, where SMC is headquartered, are also looking to reduce their massive energy consumption by promoting “green” data centers to support their AI ambitions, with the country investing more than S$500 million ($379.7 million).

Sustainable Metal Cloud is one of Asia’s largest data center operators and received funding from ST Telemedia Global Data Centres, which is backed by Singapore government investor Temasek.

SMC is currently raising $400 million in equity and $550 million in debt, which will be used to expand its data centers outside Singapore, Bloomberg reported, citing people familiar with the matter.

As the need for performance increases, so does the risk of overheating, leading technology companies to consider liquid cooling for their data centers.

Traditionally, data centers have wide aisles to circulate cooled air because overheating can cause equipment failure and downtime. In theory, liquid cooling should enable these facilities to fit many more servers within their existing footprint.

Giordano Albertazzi, CEO of digital infrastructure provider Vertiv, told CNBC in June that adoption of liquid cooling could accelerate in 2024. Vertiv’s thermal management products include hybrid air and liquid cooling, as well as fully liquid-cooled data centers.

At Computex Taipei in June, Nvidia CEO Jensen Huang touted the benefits of Supermicro’s direct liquid cooling technology, saying it would save energy and reduce costs for AI data centers.

Supermicro CEO Charles Liang told CNBC in June that liquid cooling is more power efficient, improves performance, reduces pollution and lowers energy costs.

Despite all the hype, challenges remain with adopting liquid cooling, according to SMC’s Rosenfield.

“Most data centers aren’t ready for any kind of liquid, whether it’s immersion or direct chip cooling. The market is trying to figure out the best way to utilize this, and I think there are multiple ways,” Rosenfield said.

“There’s still a lot of air cooling going on in data centers, and it will continue to be there even in full high-density AI data centers,” Vertiv’s Albertazzi said.