SuperDielectrics has moved its water-based zinc battery from a promise about safer storage into a more specific test: can a battery built around an aqueous electrolyte and patented polymer separator handle the sharp power swings that AI data centers are starting to impose on electrical infrastructure? The Cambridge, UK company says independent testing by QinetiQ showed its next-generation cell outperformed lithium-ion comparison cells in high-power cycling, charge speed, discharge speed and abuse safety. The headline figures are eye-catching: up to 13 times longer cycle life under fast charge and discharge cycling, roughly 10 times better discharge performance at room temperature, roughly 8 times better charge performance, and no thermal runaway, fire or explosion under abuse conditions. Those claims do not make the technology a direct replacement for lithium-ion in every market. They do make SuperDielectrics a company to watch in one narrow but increasingly valuable use case: short-duration, high-power buffering close to data-center racks, renewable assets and other equipment that cannot tolerate unstable power quality. AI-generated editorial illustration of aqueous zinc battery testing. What QinetiQ Tested SuperDielectrics says QinetiQ tested the company's single-layer water-based zinc cells against like-for-like lithium-ion and phosphate-based lithium-ion cells. The comparison focused on high-power applications rather than EV range or multi-hour storage. That distinction matters because data centers and power-quality systems often need fast response more than long discharge duration. At room temperature, the company reported up to 13 times longer cycle life during 10-minute charge and discharge cycles at 100 percent depth of discharge. It also reported discharge performance at 100C, equal to a 36-second discharge, while maintaining more than 85 percent nominal capacity. On charging, it reported 50C performance, equal to 1 minute and 12 seconds, while maintaining more than 70 percent nominal capacity. At 0 C, the company said charge performance was about 48 times better than the comparison cells. For operators thinking about equipment rooms, edge facilities or mixed climate installations, low-temperature behavior can decide how much thermal management hardware is needed around the battery. 13x Reported cycle-life gain under high-power cycling. 100C Reported discharge rate while retaining more than 85 percent capacity. 0 Thermal runaway events reported in abuse testing. Why This Is a Data-Center Story AI data centers are not just using more electricity. Their loads can change quickly as training clusters, inference workloads and cooling systems ramp up and down. Utilities, generators and onsite power systems have to absorb those changes without creating voltage instability, brownout risk or excess stress on backup equipment. Lithium-ion batteries can help, but they are usually optimized around energy density and longer backup windows. When they are used for repeated high-power pulses, developers may oversize packs to manage degradation, safety setbacks and warranty exposure. That adds cost, floor space, cooling equipment and siting complexity. SuperDielectrics is pitching a different layer in the power stack. Its rack-level product is being designed as a fast buffer that can sit closer to mission-critical equipment because the water-based electrolyte removes the thermal-runaway risk that makes lithium-ion harder to place inside or near high-value infrastructure. AI-generated editorial illustration of battery buffer racks inside a power-quality room. The Chemistry Tradeoff Water-based zinc batteries have a clear materials argument. Zinc is abundant, less exposed to critical-mineral bottlenecks than nickel or cobalt, and easier to explain to customers worried about supply-chain risk. Aqueous systems also offer a safety story that lithium-ion suppliers have to answer with pack design, spacing, fire suppression and operating controls. The tradeoff is energy density. SuperDielectrics is not claiming that its zinc-polymer system can replace high-energy EV cells or every grid-storage battery. Its public material points instead to fast power delivery, safety and secure supply. That makes the product closer to a power-quality and buffering asset than a classic four-hour grid battery. That may be enough. Data centers already buy uninterruptible power supplies, generators, switchgear, power conditioning and thermal systems because uptime is more valuable than hardware simplicity. If a battery can reduce overbuilding, avoid fire-safety spacing and respond quickly to power spikes, it can win a role even without matching lithium-ion on stored energy per kilogram. What Still Needs Proof Independent cell testing is not the same as commercial deployment. SuperDielectrics still has to turn lab and single-layer performance into packaged systems with controls, power electronics, safety certifications, service procedures and customer economics. Rack-level integration will be the real test because data-center operators buy uptime, not cell chemistry. The company says the results support development of its Faraday 3 battery, with a first deployment under commercial conditions planned for early next year. That timeline puts the next year of validation into focus. Customers will want operating data across repeated load swings, heat, humidity, standby behavior and failure modes. Cost will be watched just as closely. Zinc and water-based components can support a lower-cost story, but early systems often carry engineering, manufacturing and certification costs that blunt the raw-material advantage. The company needs to show total installed cost, not only cell-level performance. AI-generated editorial illustration of a data center linked to renewable power and battery buffering. Where It Fits Beside Other Storage Technologies The battery industry is splitting into more specialized categories. Lithium iron phosphate is becoming the default workhorse for many grid and EV applications. Sodium-ion is being pushed toward lower-cost stationary storage and cold-weather use cases. Iron-air, flow batteries and other long-duration systems are aimed at multi-day grid needs. SuperDielectrics is aiming at a faster, closer-to-load niche. That niche is getting more important because AI infrastructure is changing the shape of electricity demand. Operators want power that is available, clean, safe and stable. Utilities want customers to smooth their own peaks. Data-center developers want equipment that can fit inside constrained sites without raising insurance and permitting risk. Aqueous zinc will not win every storage job. It could become useful where the battery is judged by response speed, safety and cycle abuse rather than by how many hours it can discharge at rated power. The Bottom Line SuperDielectrics' QinetiQ results give water-based zinc batteries a sharper commercial target. The company is not trying to out-range EV lithium-ion cells. It is trying to solve a power-quality problem that AI data centers are making more urgent. The next proof point is not another lab number. It is a rack-level system operating under commercial conditions, with clear data on response time, safety, degradation, footprint and cost. If that system works, aqueous zinc could become part of the data-center power stack, sitting beside lithium-ion, generators and grid equipment as a fast buffer for volatile loads. Sources SuperDielectrics, June 29, 2026 TechRadar Pro, July 13, 2026 Data Centre Solutions, July 2026 Engineer Live, July 2026