Motivair Corporation’s Post

Do you have the right cooling technology to put AI to work? The average rack power density today is around 15 kW/rack, but AI workloads, according to industry reports, will command upwards of 60 – 120 kW/rack. With the accelerating pace of high-density workloads, direct #liquidcooling technology will be the bridge for #datacenters in the #AI era. Here’s why: 🌡️ Superior Thermal Performance: Liquid cooling offers superior heat dissipation capabilities compared to traditional data center cooling methods. 🔋 Greater Energy Efficiency: Single-Phase Direct Liquid Cooling uses less energy compared to other data center cooling methods. 💻 Higher Computing Density: Liquid cooling maximizes the utilization of available space. This allows data centers to accommodate more computing power in the same footprint, optimizing resources & increasing revenue-generating potential. What can our cooling technology do for you? https://lnkd.in/dwgaq-P

#AI computing is HOT! 🔥🔥🔥 Traditional air cooling in data centers will become ineffective in cooling the next generation of AI superchips. As #liquid cooling gradually becomes the industry standard, a leak protection design is required in more and more upcoming projects. We’ve designed leak sensor boards for all platforms, providing warnings when encountering coolant leakage to relieve the biggest concern of adopting liquid cooling systems. Learn more about DLC solution 👉 https://gbte.tech/gAE7 #GIGABYTE #GIGABYTEgroup #GigaComputing #serversolutions

Liquid cooling replaces air cooling in #datacentres because of the massive increase of rack densities due to #AI #sustainability #emissions #datasciences https://lnkd.in/dHSh_GCU

Simply amazing! #Coolinsights #Directliquidcooling #AcceleratedCompute #AI #ML #GenerativeAI #Artificialintelligence #Liquidcooling #Highdensity #DataCenters #HPC #CoolIT #CoolITSystems #CoolingTechnology #HighPerformanceComputing #Coolthefuture #DLC

Cool 2 x NVIDIA® GB200 NVL72 racks and other liquid-cooled AI with no facility water required. Key benefits: ➡ Highest capacity liquid-to-air CDU at 180 kW @ 15°C approach within standard data center airflow range. ➡ High 1.5 LPM / kW flow-to-heat ratio. ➡ Precise temperature flow and pressure control. ➡ Dynamic heat load management with immediate response for variable thermal loads. ➡ Low CFM and power consumption per kW of heat dissipated. ➡ Group management: up to 20 CDUs. ➡ High availability with built-in redundancies. ➡ High reliability and field serviceable design. Units are in production and shipping from CoolIT’s Canadian manufacturing facility now! Learn more: https://lnkd.in/dKAbS6Vw #Coolinsights #Directliquidcooling #AcceleratedCompute #AI #ML #GenerativeAI #Artificialintelligence #Liquidcooling #Highdensity #DataCenters #HPC #CoolIT #CoolITSystems #CoolingTechnology #HighPerformanceComputing #Coolthefuture #DLC

Interesting article in the WSJ on the need for cooling of data centers and AI processing chips. Our high and constantly growing dependency on computing power and data storage, including AI and social media (!) clearly has an impact on energy- and water use. Currently, data centers consume roughly 3% of US electricity, expected to grow to 8% by 2030, and cooling of the processing chips and data servers account for up to 40% of data center energy use. With new advanced liquid based cooling technologies, this can be reduced to 10% according to experts. Cooling clearly holds a substantial energy efficiency potential. What's not covered in the article is the high water use and the chemistries and potential negative environmental impact from the cooling liquids used. Typically these liquids are mineral oils, various fluorinated liquids or glycol/HFCs. Not very green, regardless of the spin. According to various sources, a large data center consume anywhere from 1 million to 5 million gallons of water a day - as much as what a town of 20,000 to 100,000 people use. With advanced cooling technologies, the water needs of these electricity hungry data centers can be reduced considerably. The industry's current water use trajectory is not sustainable. #waterenergynexus #datacenters #ai #wateruse

"Cooling demands of artificial intelligence and other high-density workloads have outstripped the capabilities of air systems, requiring some form of liquid cooling." At JETCOOL Technologies Inc. our goal is to drive data center efficiency by harnessing our patented microconvective liquid cooling technology for high-density applications such as AI and HPC. Our proprietary technology empowers precision cooling at the chip level, enabling more efficient and sustainable cooling systems for enterprises. A big shoutout to Sebastian Moss for featuring JetCool’s technology in the DatacenterDynamics article below, link here: https://lnkd.in/ejmZ_yhT

When it comes to AI workloads, power is becoming a bottleneck. Datacenters are ramping up to meet the demands of AI workloads. With platforms like Nvidia's Grace Blackwell, consumption is hitting 3.3 kW, and addressing both power requirements and cooling requirements is mission critical to realizing growth potential. Cooling is a core competency of Phononic. Our solutions are approved for all US hyperscaler data centers, with more than 20 million transceivers strong in field with no field issues or recalls. Phononic highly cost-competitive heat-flow technology delivers up to 30% lower power to transceiver and rack applications in datacenters—that’s nearly twice the heat pumping capability. Learn more at:https://lnkd.in/es6kMrCS #AIWorkloads #DatacenterCooling #PowerEfficiency #HyperscaleDatacenters #ThermalManagement

#AI and high-performance computing are driving the need for more advanced cooling technology. Immersion cooling could satisfy that need. Read more from Network World.

We have Ai-Ready data centers From machine learning (#ML) to generative artificial intelligence (#AI), learn how you can leverage our high-performance compute (#HPC) architecture today. - Ultra-high-density power of from 30-50 kW/rack 🚀 - Low latency connectivity to inference points, particularly for real-time immersive AI 🤖 - Advanced cooling for heat-intensive rendering and extended hardware lifespan 🆒 - 100% renewable power, with defined path to 24/7 Carbon Free Energy operations ✅ - Asset Lifecycle Management (ALM) for optimized compute performance and recycling/remarketing/resale of precious servers, CPUs and GPUs 🧭 Find out everything that sets iron mountian apart from the competition when it comes to Ai via the link http://spr.ly/6049dYTPV