EVE Energy Showcases Comprehensive BBU Solutions at OCP China 2026 to Support 800V HVDC AI Data Centers

BEIJING — At the 2026 Open Compute Project China Summit (OCP China) held here on July 9, EVE Energy made a prominent appearance with its full portfolio of Battery Backup Unit (BBU) products, positioning the lineup as the core building block of its end-to-end energy storage architecture for AI data centers. The event gathered industry leaders to discuss next-generation AI data center infrastructure, high-voltage direct current (HVDC) power delivery, and computing–energy collaboration.

Dr. Zhang Xinyi, Director of the New System Research Center at EVE Energy's Battery System Research Institute, delivered a keynote address that dissected transformative trends in AI data centers and outlined pathways for energy storage innovation. Her presentation underscored the company's strategic focus on bridging the gap between surging compute demands and resilient power infrastructure.

As large-scale deployment of AIGC model training and inference workloads accelerates, data centers are facing exponential growth in peak computational loads, millisecond-level uninterruptible power requirements, and around-the-clock renewable-energy peak shaving. EVE Energy's full-link energy storage portfolio addresses these challenges across rack, distribution, and grid-connected tiers. At the rack level, its BBU solutions smooth out high-frequency GPU power fluctuations that arise during model iteration, preventing costly training disruptions. For distribution-side applications, the "Panshield Series" 6C/10C high-rate backup cabinets operate reliably across an extreme temperature range from –30°C to 80°C, meeting short-duration backup needs with robust performance. On the source-grid side, the Mr. Big Family long-duration energy storage systems integrate with wind and solar power, leveraging the ultra-safe stacking-cell technology and extended float-charge lifespan of EVE's prismatic cells to deliver a stable power foundation for always-on AIGC clusters. This tiered approach addresses the triple challenge of compute-related energy consumption, supply security, and green-energy compliance.

Dr. Zhang noted that the continuous evolution of AI compute clusters is steering the industry toward 800V HVDC as the dominant architecture. For the unique, high-amplitude, and frequent load fluctuations inherent in AI training scenarios, EVE Energy's BBU provides dual critical value: seamless millisecond-level power takeover during grid anomalies to sustain uninterrupted compute operations, and distributed rack-level energy storage that actively dampens instantaneous power surges. This dual capability mitigates the risk of compute outages caused by grid disturbances, covering operational regimes that traditional capacitors and diesel generator sets cannot fully address.

Looking ahead to the long-term development of the intelligent computing industry, EVE Energy is advancing four core technology pillars: comprehensive cell-performance upgrades, high-density power-electronics development, a transition toward liquid-cooling thermal management, and intelligent active control of storage systems. The company's R&D teams have completed multiple rounds of system-level simulations, safety-combustion tests, and extreme-load scenario validations. Key ongoing projects include liquid-cooled BBU units and full-system safety certifications, with steady progress reported across all fronts.

"In the era of computing power, electricity is the foundation," a company representative commented. "EVE Energy ensures that every watt of compute capacity is backed by reliable energy storage." Through its full-link AIDC energy storage framework spanning source-grid, distribution, and rack levels, EVE Energy is inviting partners across the industrial chain to co-create solutions—leveraging mature self-developed products for immediate deployment, joint R&D for cutting-edge technologies, and collaborative ecosystem building for comprehensive computing–energy synergy. The company aims to jointly explore the emerging GW-level blue ocean of AI data center power–compute coordination.

Edit by paco

Last Update:2026-07-27 08:48:02