ZutaCore and Compal Transform AI Data Center Cooling at Yotta 2025

Rhea-AI Summary

Compal (OTC:CMPFF) and ZutaCore have announced a strategic partnership to revolutionize AI data center cooling solutions, showcasing their innovations at Yotta 2025. The collaboration integrates ZutaCore's HyperCool® waterless liquid cooling technology with Compal's high-performance server platforms.

Two key platforms were unveiled: the SG720-2A-L2, a 7U 8-GPU system achieving PUE of 1.04-1.1, and the SD221-8A-L2, a 2U 4-Node platform reaching an industry-leading pPUE of 1.01. The solutions utilize non-conductive, non-corrosive dielectric fluid for direct-to-chip cooling, offering enhanced safety and efficiency for AI workloads.

The technology promises higher rack performance, lower energy costs, improved reliability, and significant sustainability benefits for data center operations.

Positive

• Achieved breakthrough PUE scores of 1.04-1.1 for SG720-2A-L2 and pPUE of 1.01 for SD221-8A-L2
• Non-conductive, non-corrosive cooling technology eliminates hardware damage risk
• Dual compatibility with both EIA 19-inch racks and ORv3 infrastructure
• Technology enables higher AI workload density while reducing energy costs

Negative

• None.

Breakthrough waterless, two-phase liquid cooling delivers unmatched efficiency, reliability, and sustainability for next-generation AI workloads

LAS VEGAS, Sept. 8, 2025 /PRNewswire/ -- In response to AI's surging power demands and unprecedented computing needs, Compal, a pioneer in high-performance AI server platforms, and ZutaCore®, a leader in waterless direct-to-chip liquid cooling, have joined forces to create integrated server and cooling solutions that revolutionize data center efficiency, reliability, and sustainability. The collaboration will be on display at Yotta 2025, booth #316.

"Partnering with ZutaCore allows us to bring truly transformative cooling technology into our high-performance server portfolio," said Alan Chang, Vice President of the Infrastructure Solutions Business Unit, Compal. "The combination of our engineering expertise and ZutaCore's HyperCool® platform enables data centers to scale AI workloads more sustainably and cost-effectively."

My D. Truong, ZutaCore CTO, said the partnership delivers the best of both technologies designed for AI's most demanding workloads—from hyperscale data centers to the edge. "With Edge AI accelerating faster than ever, cooling becomes a critical enabler of real-time intelligence,' he said. "Our waterless two-phase direct-to-chip cooling uses a non-conductive, non-corrosive dielectric fluid that instantly removes heat at the source, keeping processors cool even in the most constrained edge environments. Unlike water-based systems, it is 100% IT-safe—even in the unlikely event of a leak, it will not harm hardware. This breakthrough allows operators to deploy AI closer to where data is generated, while maximizing performance per rack, slashing energy costs, and advancing sustainability goals without compromising reliability."

From Innovation to Implementation: Next-Generation Platforms

At Yotta 2025, attendees will see two Compal platforms built with ZutaCore's HyperCool® waterless liquid cooling, proving how advanced cooling translates into real-world gains in density, efficiency, performance and reliability.

• SG720-2A-L2 (7U 8-GPU UBB)
  Designed for next-generation AI acceleration, this system features AMD Instinct™ MI325X GPUs and is ready for the upcoming MI355X. Its 850mm chassis delivers maximum flexibility: it fits seamlessly into today's EIA 19-inch racks—the universal rack size used in nearly every data center—while also supporting ORv3 (Open Rack Version 3) infrastructure, a new open standard from the Open Compute Project backed by Meta and Microsoft that enables hyperscale efficiency and sustainability. Supporting both ensures compatibility with today's environments and future-readiness for tomorrow's. A modular, tool-less design streamlines maintenance and upgrades, while liquid cooling integration enables an impressive PUE (Power Usage Effectiveness) score as low as 1.04–1.1
• SD221-8A-L2 (2U 4-Node Dual Processor)
  Built for high-density computing, this platform redefines efficiency with a 2U4N architecture and waterless two-phase cooling, achieving an industry-leading pPUE (partial Power Usage Effectiveness) of 1.01. pPUE measures efficiency at the rack level, showing how little extra energy is needed beyond the servers themselves. Unlike water-based systems, ZutaCore's cooling uses a non-conductive, non-corrosive dielectric heat transfer fluid that is completely IT-safe, protecting hardware even in the event of a leak. This design minimizes risk, ensures reliability under heavy AI workloads, and combines sustainability with peace of mind.

"PUE and pPUE are the industry's gold standards for energy efficiency," noted Truong. "A perfect score of 1.0 means every watt of power goes into computing, with none wasted on cooling. Reaching values this close to 1.0 is a breakthrough that redefines sustainable, high-performance data center operations."

ZutaCore's Cooling Technology Brings Game-Changing Benefits

ZutaCore's waterless, direct-to-chip liquid cooling, with its exclusive pool boiling technology, represents a fundamental shift from traditional air or single-phase liquid cooling. Instead of moving heat away gradually, it removes it instantly at the source by turning liquid into vapor, keeping processors consistently cool even under the most extreme workloads.

For operators, the impact is immediate:

• Higher performance per rack — Run more AI and HPC workloads in the same footprint.
• Lower energy costs — Drive PUE and pPUE closer to the ideal of 1.0, cutting power bills dramatically.
• Greater reliability — Eliminate risks of condensation, corrosion, or hardware damage.
• Sustainability gains — Reduce both water and electricity use, accelerating ESG progress.

About Compal

Founded in 1984, Compal is a leading manufacturer in the notebook and smart device industry, creating brand value in collaboration with various sectors. Its groundbreaking product designs have received numerous international awards. In 2025, Compal was recognized by CommonWealth Magazine as one of Taiwan's top 7 manufacturers and has consistently ranked among the Forbes Global 2000 companies. In recent years, Compal has actively developed emerging businesses, including cloud servers, auto electronics, and smart medical, leveraging its integrated hardware and software R&D and manufacturing capabilities to create relevant solutions. More information, please visit https://www.compal.com

View original content to download multimedia:https://www.prnewswire.com/news-releases/zutacore-and-compal-transform-ai-data-center-cooling-at-yotta-2025-302548739.html

SOURCE COMPAL ELECTRONICS INC.

FAQ

What are the key features of Compal and ZutaCore's new cooling solution for AI data centers?

The solution features waterless, two-phase liquid cooling using non-conductive dielectric fluid, achieving PUE scores as low as 1.04-1.1 for the SG720-2A-L2 and pPUE of 1.01 for the SD221-8A-L2 platform.

How does ZutaCore's cooling technology differ from traditional data center cooling methods?

ZutaCore uses waterless direct-to-chip liquid cooling with pool boiling technology that instantly removes heat by turning liquid into vapor, unlike traditional air or single-phase liquid cooling systems.

What efficiency improvements does Compal's new server platform achieve with ZutaCore cooling?

The SD221-8A-L2 platform achieves an industry-leading pPUE of 1.01, while the SG720-2A-L2 system reaches PUE scores of 1.04-1.1, significantly improving energy efficiency.

What are the main benefits of the Compal-ZutaCore cooling solution for data center operators?

The solution offers higher performance per rack, lower energy costs, improved reliability with no risk of hardware damage, and enhanced sustainability through reduced water and electricity usage.

How does the new cooling technology address AI workload challenges?

The technology enables instant heat removal at the source, maintaining consistent processor cooling even under extreme AI workloads, while allowing for higher density computing and improved energy efficiency.