伙伴关系
EPRI的DCFlex倡议正式成员
在 Google 上的首选来源中添加 Unite.AI
Emerald AI 宣布于2026年11月2日表示，它正作为正式成员加入EPRI的DCFlex Initiative，合作范围包括AI数据中心灵活性的六个演示
Emerald AI 称其与EPRI的关系日期为2024年11月公司成立，同年秋季 DCFlex 发布。 示威涉及盐河项目，Oracle,ComEd,PJM互联，星座，多米尼翁能源，波特兰通用电气，国家网格和内比乌斯。 据该公司称，NVIDIA参加了每场比赛，而翡翠导师则协调AI工作量以应对电网需求。
EPRI总裁兼CEO Arshad Mansoor欢迎翡翠AI成为正式成员。 称现场演示旨在提供一个框架，在支持电网稳定和AI驱动增长的同时，将数据中心能源使用灵活性投入运行。
报告的示范结果
Emerald AI将其方法描述为允许数据中心在有限的电网压力期间调整电力消耗，同时保护关键计算工作量的性能。 广泛应用这种灵活性的称公司可以在现有的美国电力系统上解锁多达100GW的功率，支持更快和更大的AI基础设施连接，同时帮助维持可靠，负担得起的电力。
据Emerald AI称，亚利桑那州演示结果发表在《自然能源》上，在随后的一次伦敦示威中，称翡翠操控公司达到了200多个电力目标，削减率达到40%.
另一场演示改变了弗吉尼亚州和芝加哥之间的推论工作量，以缓解当地力量的制约，同时保持低潜。 在俄勒冈州，翡翠AI与NVIDIA的全DSX堆叠融合，并在包括冻天气和长期热浪，称公司在内的情景下响应了公用事业请求。
研究计划和DSX工作
创始人兼首席执行官Varun Sivaram 称 Emerald AI正致力于建立一种模式，通过为电网提供可靠的灵活性，AI数据中心能够更快地确保电力。 他描述公司的雄心壮志是支持AI增长，同时加强电力系统，保护该基础设施所在社区的电力承受能力。
通过正式的DCFlex成员，Emerald AI 称将与EPRI合作进行独立研究。 称公用事业公司需要清楚了解一个设施能够提供的反应，包括其规模、速度、持续时间和在特定电网条件下的可用性。 EPRI的示范和共享技术框架有助于为评估这些能力建立一致的基础。 对Emerald AI来说，更深入的参与旨在帮助使灵活性成为一种可靠的资源，公用事业部门可以将这种资源纳入规划和业务决定。
Emerald AI和NVIDIA也在合作，将电源灵活性纳入AI工厂的DSX参考架构，包括在NVIDIA的AI工厂研究中心的商业规模。 将灵活性纳入设施设计的翡翠AI 称为审查计算性能、设施运行和电网反应之间的关系创造了机会。 该公司认为，研究结果可以为今后的设施设计提供信息，以及如何在互联和运行期间评估灵活性。
相关政策工作
Emerald AI 称其 DCFlex 参与补充了其作为AI能源管理联盟(AEMA)创始成员Google和NVIDIA的作用。 该联盟倡导采取政策，承认经核实的灵活性，支持加快数据中心的互联，以提高电网可靠性和能源承受能力。
称公司通过DCFlex和AEMA的政策工作开发的技术证据针对更广泛的采用的不同要求。 Emerald AI 称在转向更大的部署时，其重点仍然是促进AI基础设施的增长，使公用事业能够充满信心地容纳，并使同一电力系统服务的社区受益。
本文由 AI图鉴自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
Partnerships
Emerald AI Becomes Full Member of EPRI’s DCFlex Initiative
Add Unite.AI to your preferred sources on Google
Emerald AI announced on November 2, 2026 that it is joining EPRI’s DCFlex Initiative as a full member, extending a collaboration that has included six demonstrations of AI data-center flexibility with NVIDIA, utilities, grid operators and data-center partners.
Emerald AI said its relationship with EPRI dates to the company’s founding in November 2024, the same fall DCFlex launched. The demonstrations have involved Salt River Project, Oracle, ComEd, PJM Interconnection, Constellation, Dominion Energy, Portland General Electric, National Grid and Nebius. According to the company, NVIDIA participated in each one, while Emerald Conductor coordinated AI workloads in response to grid needs.
EPRI President and CEO Arshad Mansoor welcomed Emerald AI as a full member. He said the field demonstrations were intended to inform a framework for putting data-center energy-use flexibility into operation while supporting grid stability and AI-driven growth.
Reported Demonstration Results
Emerald AI describes its approach as allowing data centers to adjust electricity consumption during limited periods of grid stress while protecting the performance of critical computing workloads. The company said applying that flexibility broadly could unlock up to 100 GW of capacity on the existing U.S. power system, supporting faster and larger AI infrastructure connections while helping maintain reliable, affordable electricity.
Results from the Arizona demonstration, published in Nature Energy, showed an AI computing cluster reducing power draw by 25% for three hours during peak demand while meeting all workload service commitments, according to Emerald AI. In a subsequent London demonstration, the company said Emerald Conductor met more than 200 power targets, with reductions reaching 40%.
Another demonstration shifted inference workloads between Virginia and Chicago to relieve a local power constraint while maintaining low latency. In Oregon, Emerald AI integrated with NVIDIA’s full DSX stack and responded to utility requests under scenarios including freezing weather and a prolonged heat wave, the company said.
Research Plans and DSX Work
Founder and CEO Varun Sivaram said Emerald AI is working toward a model in which AI data centers can secure power sooner by providing dependable flexibility to the grid. He described the company’s ambition as supporting AI growth while strengthening the power system and protecting electricity affordability in the communities hosting that infrastructure.
Through full DCFlex membership, Emerald AI said it will collaborate with EPRI on independent research. The company said utilities need a clear understanding of the response a facility can provide, including its magnitude, speed, duration and availability under specific grid conditions. EPRI’s demonstrations and shared technical frameworks help establish a consistent basis for evaluating those capabilities. For Emerald AI, deeper participation is intended to help make flexibility a dependable resource that utilities can incorporate into planning and operating decisions.
Emerald AI and NVIDIA are also collaborating to integrate power flexibility into the DSX reference architecture for AI Factories, including at commercial scale at NVIDIA’s AI Factory Research Center. Emerald AI said incorporating flexibility into a facility’s design creates an opportunity to examine the relationships among computing performance, facility operations and grid response. According to the company, the findings can inform future facility designs and how flexibility is assessed during interconnection and operation.
Related Policy Work
Emerald AI said its DCFlex participation complements its role with Google and NVIDIA as a founding member of the AI Energy Management Alliance (AEMA). The alliance advocates policies that recognize verified flexibility and support faster interconnection for data centers that contribute to grid reliability and energy affordability.
The company said the technical evidence developed through DCFlex and AEMA’s policy work address different requirements for wider adoption. As it moves into larger deployments, Emerald AI said its focus remains on enabling AI infrastructure growth that utilities can confidently accommodate and that benefits communities served by the same power system.
Auto-fetched and translated by AI Atlas. All rights belong to the original author.