谷歌披露电网互动型AI数据中心能源与设施技术路线

Inside a Google New Albany data center buildings in Central Ohio

Inside one of Google's New Albany data center buildings in Central Ohio.Google位于俄亥俄州中部的 Google 新奥尔巴尼数据中心其中一栋建筑内。Google

AI infrastructure is expanding at an unprecedented pace. Google now processes more than 3 quadrillion AI tokens per month, representing a sevenfold increase in compute demand from May 2025 to May 2026. Meeting this velocity requires innovation in energy efficiency and power delivery.AI 基础设施正以前所未有的速度扩张。Google 目前每月处理超过 3 quadrillion 个 AI token，相当于 2025 年 5 月至 2026 年 5 月算力需求增长至七倍。要跟上这种速度，必须在能源效率和电力输送方面创新。

Tom Garvens, vice president of advanced technology innovation at Google, outlined this challenge during a keynote at Data Center World Power 2026 in Houston, held Sept. 21-23. He described Google’s full-stack approach to scaling AI, positioning data centers as active partners in stabilizing local grids.Google 先进技术创新副总裁 Tom Garvens 在 9 月 21 日至 23 日于休斯敦举行的 Data Center World Power 2026 上发表主题演讲时概述了这一挑战。他介绍了 Google 扩展 AI 的全栈方法，将数据中心定位为稳定本地电网的积极合作伙伴。

Turning Backup Power into Grid Services将备用电力转化为电网服务

To bridge the speed mismatch between rapid AI rollout and multiyear timelines typical for utility capacity expansion, Google has developed energy strategies that transform backup assets into grid-supportive resources, streamline operations during peak demand, make load shaping more intelligent, and prevent disruptive power swings – while supporting utility resilience.为弥合 AI 快速部署与公用事业容量扩建通常所需的多年周期之间的速度失配，Google 制定了能源战略：将备用资产转化为支持电网的资源，在需求高峰期简化运营，使负荷整形更加智能，并防止破坏性电力波动——同时支持公用事业韧性。

Beyond these strategies, Google is reviewing and redesigning the entire AI data center ecosystem – from chip to utility. This includes optimizing for liquid cooling, low-voltage direct current (LVDC) architectures, intelligent algorithms, and facility design. A core pillar of this work is publishing standardized power and facility technology specifications through communities such as the Open Compute Project (OCP), enabling the industry to build interoperable, efficient, and sustainable infrastructure.除这些战略外，Google 正在审视并重新设计从芯片到公用事业的整个 AI 数据中心生态系统。这包括针对液冷、低压直流（LVDC）架构、智能算法和设施设计进行优化。这项工作的核心支柱之一是通过各类社区发布标准化电力与设施技术规范，例如 Open Compute Project (OCP)，使行业能够构建可互操作、高效且可持续的基础设施。

“Google has tightly coupled vertical integration in AI along with close partnerships with energy providers,” Garvens said. “We develop and design our own data centers, chips, servers, and rack designs. By having control of our entire hardware and software stack, we can optimize these things together for sustainability and energy efficiency.”“Google 既在 AI 领域紧密推进垂直整合，又与能源供应商保持密切合作伙伴关系，”Garvens 说。“我们开发并设计自己的数据中心、芯片、服务器和机架设计。通过掌控整个硬件和软件堆栈，我们可以共同优化这些环节，以提升可持续性和能源效率。”

Powering AI at Gigawatt Scale以吉瓦级规模为 AI 供电

Google is investing alongside utilities to add grid capacity. That assistance is vital as a gigawatt-class data center campus can comprise hundreds of thousands of interconnected chips running a single workload. Google’s approach is twofold: work with utility partners to streamline power delivery and adjust its own hardware and software to be more grid-friendly.Google 正与公用事业公司共同投资增加电网容量。这种协助至关重要，因为一个吉瓦级数据中心园区可由数十万个运行单一工作负载的互联芯片组成。Google 的方法分为两方面：与公用事业合作伙伴合作简化电力输送，并调整自身硬件和软件，使其对电网更友好。

“Utilities are [setting] different requirements that we must demonstrate that we can meet,” Garvens noted. “Google, Meta, Microsoft, and Nvidia have come together to define an interface spec on how utilities and hyperscalers can interface with each other.”“公用事业公司正在[设定]不同要求，我们必须证明自己能够满足这些要求，”Garvens 指出。“Google、Meta、Microsoft 和 Nvidia 已共同定义一项接口规范，说明公用事业公司与超大规模运营商应如何相互对接。”

Battery energy storage systems (BESS) have become critical in Google’s campus designs for onsite energy storage and for buffering rapidly fluctuating workloads. Buffering smooths power and voltage transitions. Backup generators remain in use, but diesel does not scale well to campuses at 1 GW and beyond. BESS are increasingly used to provide standby power, supported by demand-response agreements with utilities for peak shaving.电池储能系统（BESS）已成为 Google 园区设计中的关键，用于现场储能以及缓冲快速波动的工作负载。缓冲可平滑功率和电压过渡。备用发电机仍在使用，但柴油发电无法很好地扩展至 1 GW 及以上规模的园区。BESS 正越来越多地用于提供备用电力，并通过与公用事业公司签订的需求响应协议为削峰提供支持。

“When peak needs are beyond what we've got power for, we can get that from local storage,” Garvens explained.“当峰值需求超出我们的供电能力时，我们可以从本地储能获取电力，”Garvens 解释道。

Rather than competing for constrained capacity in saturated markets, Google has been identifying remote sites with ample land and available power. The company is also considering campus developments in small towns where the data center would be a primary employer, he said.谷歌没有在饱和市场中争夺紧张的容量，而是一直在寻找土地充足、电力可获得的偏远场地。该公司还在考虑在小城镇开发园区，使数据中心成为主要雇主，他说。

Another design change: Google has moved away from centralized Uninterruptible Power Supply (UPS) systems, favoring rack-level battery backup operating around 54 VDC. Eliminating an extra AC-DC-AC conversion step yields approximately a 1.5% efficiency gain.另一项设计变化是：谷歌已放弃集中式不间断电源（UPS）系统，转而采用工作在约 54 VDC 的机架级电池备用电源。省去额外的AC-DC-AC 转换环节可将效率提高约 1.5%。

“We do all of our power conversion at the top of our racks, where we take our AC down to 54 volts, and then we run DC through our racks; our battery backup also runs at that voltage,” Garvens said.“我们在机架顶部完成所有电力转换，将 AC 降至 54 伏，然后通过机架输送 DC；我们的电池备用电源也在该电压下运行，”Garvens 说。

Inside the Rack: 54 VDC Today, 800 VDC Tomorrow机架内部：如今 54 VDC，未来 800 VDC

Google has introduced changes at the chip, rack, and facility levels, supported by software that can intermix workloads to smooth overall load and make it more predictable.谷歌在芯片、机架和设施层面都进行了调整，配套软件能够混合运行工作负载，以平滑总体负载并使其更可预测。

Rack density remains a major challenge. Roadmaps anticipate racks approaching 1 MW each over the next year or two. That necessitates liquid cooling for at least the hotter elements within the racks.机架密度仍是一项重大挑战。路线图预计，未来一两年每个机架将接近 1 MW。这使得液冷成为必需，至少对机架内更热的部件而言是如此。

Delivering the required current at lower voltages would demand bus bars and cabling with significantly more copper. This calls for a higher distribution voltage. Google is pursuing a gradual migration toward 800 VDC for power distribution, which is expected to take a couple of years to mature.在较低电压下输送所需电流，要求母线排和电缆使用显著更多的铜。这就需要更高的配电电压。谷歌正逐步转向采用 800 VDC 配电，预计需要几年时间才能成熟。

In the meantime, Google has implemented “sidecars” alongside its racks – adjacent power-distribution enclosures that house cabling and power gear – reserving the racks for high-density servers. The tradeoff is floor space. When 800 VDC becomes broadly available, those sidecars can be converted to additional AI server racks.与此同时，谷歌在机架旁部署了“边车”——紧邻机架的配电柜，用于容纳布线和电力设备——从而将机架留给高密度服务器。其代价是占用楼面空间。当 800 VDC 得到广泛应用时，这些边车可以改造成额外的 AI 服务器机架。

“We're working with the industry on bringing solid-state transformers to maturity,” Garvens said. “We are also continuing to improve our power usage effectiveness [PUE]. We can't ever get to 1.0, but we continue to challenge ourselves to get as close to it as possible.”“我们正与行业合作，推动固态变压器走向成熟，”Garvens 说。“我们也在持续改进电能使用效率 [PUE]。我们永远无法达到 1.0，但我们继续挑战自我，尽可能接近这一数值。”

Workload Shifting and Water-Smart Cooling工作负载转移与节水型冷却

With dozens of data centers in operation and many more in development, Google can be fluid about where it runs workloads. Garvens offered an example:谷歌已有数十个数据中心投入运营，还有更多处于开发阶段，因此可以灵活安排运行工作负载的地点。Garvens 举了一个例子：

“If there's a heat wave in Ohio and we were running a big workload, we've got agreements in place whereby with the right amount of foresight we can shut certain workloads down or move them elsewhere if they're critical,” he said. “We have a portfolio where we can shift things around.”“如果俄亥俄州出现热浪，而我们正在运行大型工作负载，我们已经签订协议；只要有适当的提前预判，我们就可以关闭某些工作负载，或者在这些工作负载至关重要时将其转移到其他地方，”他说。“我们拥有可灵活调配的资源组合。”

Water strategy is another focus area. The company is investing in closed-loop cooling systems – sealed liquid circuits where the working fluid remains within the system and makeup water is minimized. This marks a shift from designs that consumed large quantities of water. There is no single approach; depending on conditions – altitude, humidity, water table, drought risk, and community needs – Google works with partners to determine the appropriate power capacity and the best cooling design.水资源战略是另一个重点领域。该公司正在投资闭环冷却系统——密封液体回路中的工作流体保留在系统内，并将补充水量降至最低。这标志着设计思路已从耗水量大的方案发生转变。没有单一方案；根据海拔、湿度、地下水位、干旱风险和社区需求等条件，谷歌与合作伙伴共同确定合适的电力容量和最佳冷却设计。

“We work with our partners in terms of what's available and figure out what is the right amount of power we can get and what water usage or closed-loop design is most appropriate,” Garvens noted.“我们根据可用条件与合作伙伴合作，确定我们能够获得的合适功率，以及哪种用水量或闭环设计最合适，”Garvens 指出。

Keeping Pace with Silicon Roadmaps与硅技术路线图同步

Each new generation of silicon tends to draw more power, and multiple chips are arriving every year.每一代新硅产品往往会消耗更多功率，而且每年都会有多种芯片问世。

“The most precious thing we have is a watt: you don't want to ever strand a watt,” Garvens said. “So, making our data center and our designs as crisp as we can for as far out as we can see in these roadmaps, it's why our partnerships with our silicon and GPU design teams are so critical.”“我们最宝贵的东西是一瓦电：你绝不想浪费一瓦，”Garvens 说。“因此，在这些路线图可预见的范围内，让我们的数据中心和设计尽可能精简高效，这就是我们与硅芯片和 GPU 设计团队的合作如此关键的原因。”

About the Author关于作者

Contributor撰稿人

Drew Robb has been a full-time professional writer and editor for more than twenty years. He currently works freelance for a number of IT publications.Drew Robb 担任全职专业作家兼编辑已有二十多年。他目前为多家 IT 出版物从事自由撰稿工作。