\u201cI am sure people have heard about the huge investments being made in AI data\u00a0centers, but few understand the massive amount of power they require,\u201d opens \u201cIn a typical data\u00a0center, for every ten units of power used to run services, another six units of power\u00a0are required to\u00a0cool them with air. By deploying immersion, those same servers can be cooled with less than one unit of power, providing customers with two options: capture the operational cost savings of a significantly lower electric bill, or reallocate that power saving to\u00a0additional\u00a0computing, effectively achieving more productivity using the same power envelope. Also, unlike air cooling, because the servers are immersed in the coolant, they\u00a0aren\u2019t\u00a0exposed to airborne particulates, humidity, and oxidation, which improves the long-term reliability of the servers.<\/span>\u00a0<\/span><\/p>\n \u201cIn recent years, technology markets have increasingly addressed the ever-growing carbon footprint of data\u00a0centers, with large hyperscale operators like Google, Meta, and AWS making some form of carbon reduction pledge,\u201d he adds. \u201cDue to the cost of sustainable technologies, business leaders have historically had to choose whether to invest in more sustainable technologies at a higher cost and risk lower profitability or\u00a0optimize\u00a0shareholder value with conventional technologies. However, immersion eliminates that trade off, as\u00a0it\u2019s\u00a0a more sustainable technology that reduces both\u00a0CapEx\u00a0and\u00a0OpEx\u00a0while also reducing carbon footprint. There is also the benefit of reducing waste by extending the life of the servers and preventing them from going to landfill.\u201d<\/span>\u00a0<\/span><\/p>\n Narrating the story of GRC\u2019s evolution, Peter continues: \u201cWe think about the company\u2019s growth in four phases: product validation, product optimization, building a foundation for scale, and finally, growth at scale. Product validation meant deploying our\u00a0new technology\u00a0to customers like the NSA, Orange Telecom, and Intel. This successfully\u00a0demonstrated\u00a0the value of our proposition and gave us license to enter the second phase, product optimization. In this phase we sold products to early adopters, including universities, cryptocurrency miners, and oil and gas services, which then actively engaged in recommending product enhancements and helping us further\u00a0optimize\u00a0the solution. We then focused on building a foundation for scale, including our Flex Manufacturing Network, securing ISO 9001 certification, more robust information management systems, and a world-wide service and support capability. With that foundation in place, we are currently in the final phase, growth at scale, where\u00a0we\u2019re\u00a0focused on mainstream adoption of our technology.\u201d<\/span>\u00a0<\/span><\/p>\n GRC\u2019s products are composed of three basic building blocks: the coolant filled tanks into which the servers are immersed, a coolant distribution unit (CDU) that circulates the coolant to capture and reject heat from the servers out of the building, and an operational control system that also provides predictive failure alerting to avoid any unscheduled downtime. \u201cThese three components come in a range of sizes and shapes that are optimized for specific uses cases,\u201d Peter says. \u201cThe system could be as small as something you might deploy in a data closet in a branch bank or as large as systems that would be deployed in an AI data\u00a0center\u00a0the size of two football fields. The breadth of our product portfolio enables us to address applications from small, edge computing requirements to\u00a0very large, centralized cloud computing applications.<\/span>\u00a0<\/span><\/p>\n \u201cHowever, cooling is only one\u00a0component\u00a0of a data\u00a0center, and we have built an entire ecosystem of strategic partners including some of the most trusted brands in the data\u00a0center\u00a0and telecoms industries. We partner with server providers, such as a BCA with Dell, an Engineering Alliance with Cisco, and a pre-certified bundle with\u00a0SuperMicro, as well as Schnider Electric for data\u00a0center\u00a0infrastructure and several coolant providers including Shell, bp Castrol, HF Sinclair, and ENEOS to name a few.\u00a0We\u2019ve\u00a0also signed a Business Collaboration Agreement (BCA) with Samsung C&T to develop reference designs for immersion cooled data\u00a0centers\u00a0to accelerate mainstream adoption.\u201d<\/span>\u00a0<\/span><\/p>\n![\"a](\"https:\/\/manufacturing-today.com\/wp-content\/uploads\/sites\/4\/2025\/12\/GRC-243-body-image-1.jpg\")
Peter Poulin, CEO. \u201cA recent report from McKinsey projects that AI data\u00a0centers\u00a0could consume up to 100 GW of power globally, but this amount of power is simply not available in today\u2019s world. The biggest constraint to AI data\u00a0center\u00a0builds is therefore access to power, and\u00a0that\u2019s\u00a0where our immersion solution shines, by reducing power consumption by up to 40 percent.<\/span>\u00a0<\/span><\/p>\n![\"Peter](\"https:\/\/manufacturing-today.com\/wp-content\/uploads\/sites\/4\/2025\/12\/GRC-243-body-image-2.jpg\")