A Chip Company Releases AIDC Energy Storage Certification Standards. Why NVIDIA? Computing Power Reshapes Power Supply Logic. Who's in the Lead and Who's Left Out?

marsbitPublicado a 2026-06-23Actualizado a 2026-06-23

Resumen

NVIDIA has released a "Battery Energy Storage System Self-Certification Guide," setting strict technical standards for energy storage systems specifically for AI data centers (AIDC). The guide focuses solely on certifying the Power Conversion System (PCS), not the batteries, with 10 mandatory performance metrics and 12 validation tests requiring real-world and simulation comparisons. Key requirements include rapid dynamic response to AI workloads, high-frequency system telemetry, and detailed electromagnetic transient models. The move is driven by the extreme and fluctuating power demands of next-generation AI hardware. Modern AIDCs require energy storage systems to act as intelligent, controllable grid assets, not just passive backup, to manage instantaneous, massive power load shifts that traditional UPS systems cannot handle. This redefines the competitive landscape for energy storage providers, shifting focus from capacity and cost to advanced control capabilities and system integration. While the market potential is significant—with forecasts of hundreds of GWh in new demand by 2030—the certification creates a high barrier to entry. It requires proven PCS delivery volumes and credible plans for rapid capacity scaling, favoring established, well-resourced players. Early movers like Fluence (partnering with Siemens) and several Chinese companies have secured projects ahead of the standard, but new entrants must now navigate this rigorous, costly, and time-intensive certi...

Foresee Energy learned that NVIDIA recently released a "Self-Certification Guide for Energy Storage Systems," which sets 10 mandatory metrics, 12 items requiring actual measurements plus simulation comparisons, with measurement accuracy strictly defined at voltage ±0.2% and current ±0.2%. A graphics card company is drawing the entry line for the energy storage industry.

Most people's initial reaction is probably: You don't even manufacture energy storage equipment, what gives you the right to set the rules?

But in reality, as NVIDIA is redefining computing power, it's also redefining how data centers consume electricity.

From the GB200 in 2025 which first introduced the 800V HVDC architecture, to the Vera Rubin NVL72 in 2026 with a single cabinet power consumption approaching 225kW. In the Agentic AI era, thousands of GPUs within an AIDC can ramp from 10% to 100% power in milliseconds. For a 100MW AIDC, this could cause grid load to surge by tens of thousands of kilowatts instantaneously. Traditional UPS systems and diesel generators simply cannot keep up. While energy storage is being integrated into the top-level design of AI data centers, the method of its integration—through certification thresholds set by a chip company—leaves the entire industry both excited and uneasy.

01

Certification Boundary Drawn on the AC Side

It Tests the PCS, Not the Battery

The uniqueness of NVIDIA's guide lies in the fact that it focuses solely on the PCS (Power Conversion System). Battery capacity? Irrelevant. DC-side topology? Irrelevant. Battery cell type? Also irrelevant. The certification boundary is set on the AC side, with the PCS being the sole object of assessment.

Among the 10 mandatory metrics, "AI Buffer Dynamic Response" requires the system to avoid oscillations or control "chasing" during rapid power conversion; "Telemetry & Control" demands polling of all nodes at a 1Hz frequency, supporting 3 concurrent Modbus TCP connections; "Control Transparency" requires providing EMT (Electromagnetic Transient) models and impedance/admittance scan verification, and compliance with NERC reliability guidelines. All 12 items require both hardware measurements and simulation comparisons before submission for certification. Beyond the technical threshold, manufacturers must also submit their PCS delivery volume over the past 12 months and a viable plan to achieve a 10x production expansion within 24 months.

In a technical blog, NVIDIA stated: A BESS (Battery Energy Storage System) is an "intelligent, controllable electrical asset," not a passive storage warehouse. The complexity of a BESS in an AIDC factory far exceeds capacity rating—it's a control system deeply interacting with the grid, requiring full-stack co-design of hardware and software, rather than "determining capacity first, then matching control." Hardware specs alone cannot solve control issues; rapid telemetry, real-time analytics, and coordinated control architecture are key to the design. No matter how large the battery, poor control logic means failure to pass certification.

This means the competition dimensions the energy storage industry has focused on in recent years—capacity and cost—become invalid under this standard.

02

The Game Has Changed

Those Who Jumped the Gun Are Already on the Road

In early June, Siemens officially released its reference electrical and power architecture design for the NVIDIA DSX Vera Rubin NVL72 platform. Fluence's SmartStack battery energy storage system was incorporated, becoming the only explicitly designated battery storage partner in this reference design, which was met with a 43.8% stock surge on the same day. This reference design has a total facility capacity of 136MW, with IT load at 100MW. The SmartStack energy storage system configured by Fluence is 120MW/240MWh.

However, Fluence securing this position is less a victory in a technical bid and more a case of pre-emptive locking of industry chain influence. Fluence itself is a joint venture between Siemens and the American power company AES. Being included in the parent company's official reference design is essentially "niche inheritance." Other independent energy storage companies find it difficult to replicate this structural advantage.

With the release of NVIDIA's guide, the path has become clearer, but also dauntingly narrow. The market potential is not small, but are the competitors really enough? CLSA estimates that AIDC construction in China over the next five years will bring 125GWh of new demand for energy storage batteries. Guosheng Securities forecasts that just AIDC deployments in the US from 2026 to 2028 could drive incremental demand of 10GWh, 27GWh, and 39GWh, respectively. Morgan Stanley predicts that by 2030, AI data centers will generate 321GWh of annual new energy storage demand.

Some are already racing ahead. In February of this year, Nanjing Guanlong Power won a bid for a backup power project at an NVIDIA data center in Asia, supplying two sets of 1MW grid-forming energy storage converter integrated systems to replace diesel generators, enabling zero-delay seamless backup during power loss. WeiGuang Energy signed an agreement for the Changzhou Xinbei District, planning to build a solid-state transformer production base for AIDC scenarios over the next three years. Its "Xihe 2.0" system directly converts 10kV medium voltage to 800V DC, achieving a system efficiency of 98.6% and a volume only one-third that of a traditional UPS. According to Eastmoney.com, Sungrow Power had secured over 11GWh of AIDC energy storage orders by the first half of 2026. Trina Solar's energy storage business grew by over 300% year-on-year in Q1 2026, with overseas business accounting for over 90% of the total. In April, Far East Smarter Energy's orders involving smart battery storage and computing power/AI amounted to a combined 7.37 billion yuan.

However, all these projects broke ground before the formal release of the certification standards. After the standards are out, latecomers must follow the new rules from scratch.

03

The Threshold is Set

Who Can Pass, and Who Cannot?

The most brutal part of the certification guide is hidden in its final pages. Manufacturers must submit their PCS delivery volume from the past 12 months and a credible plan for 10x production expansion within 24 months. These two requirements alone shut out small manufacturers—they can't even gather the materials for certification, let alone provide delivery records or expansion commitments.

AIDC power supply energy storage systems involve multi-dimensional testing covering electrical safety, thermal runaway protection, and cybersecurity. The certification cycle, capital costs, and technical rectification difficulty are incomparable to those of typical commercial & industrial energy storage. A complete AIDC integrated energy solution isn't a single device certification; it's a combined verification of energy storage batteries, converters, power distribution equipment, and monitoring systems. It requires simultaneous completion of dozens of standard tests like UL1973 for battery cell safety, UL9540A for cabinet-level thermal runaway propagation, and IEC62443 for industrial network protection. The entire process typically takes 12 to 24 months, with cumulative costs for testing, rectification, and certification services reaching millions of RMB.

Price wars in the energy storage industry have escalated from competition at the "fen" level to the "li" level. NVIDIA's guide essentially draws a new starting line. Those lacking sufficient control capability don't even get a seat at the table. At the 2026 Two Sessions, National People's Congress representative Xu Yanming pointed out that AIDCs require energy storage systems to possess instantaneous high-current output capability and millisecond-level response speed. However, existing energy storage products still struggle to fully meet the high dynamic load demands in terms of rate performance, dynamic response, and long-term operational stability.

NVIDIA isn't trying to do the work of an energy storage association; it's drawing blueprints for its own factories. Whoever defines the computing power requirements has the right to define the power supply standards. This door has been opened, but whether one can step through depends on whether each company's PCS control algorithms are fast enough, their delivery records solid enough, and their expansion plans credible enough. Piling batteries higher won't solve this problem.

This article is from the WeChat public account "Foresee Energy," author: Zhao Jianan.

Criptos en tendencia

Preguntas relacionadas

QWhat is the core focus of NVIDIA's 'Self-Certification Guide for Energy Storage Systems' for AIDC?

AThe guide focuses exclusively on certifying the Power Conversion System (PCS), not the batteries themselves. It sets stringent performance and control requirements for PCS, such as dynamic response, telemetry frequency, control transparency, and the need to provide historical delivery data and scalable production plans.

QWhy is NVIDIA, a chip company, setting certification standards for energy storage in data centers?

ABecause NVIDIA's high-power AI chips (like GB200, Vera Rubin NVL72) are redefining power consumption in data centers, causing extreme, instantaneous power load surges. To ensure the reliability of its AI data centers (AIDC), NVIDIA is defining the power supply standards. It dictates the energy storage performance requirements because it designs the primary power-consuming systems.

QAccording to the article, how does NVIDIA's standard change the competitive landscape for energy storage companies?

AIt shifts the competition from a focus on scale, capacity, and cost to a focus on high-level control technology, software-hardware integration, and proven delivery capabilities. Companies with advanced PCS control algorithms, strong delivery records, and credible expansion plans have an advantage. Smaller companies without these capabilities may be excluded.

QWhich company was highlighted as an early beneficiary of the trend towards AIDC energy storage, and why?

AFluence (a Siemens & AES joint venture) was highlighted. Its SmartStack BESS was designated as the sole energy storage partner in Siemens's official reference design for NVIDIA's Vera Rubin NVL72 platform. This advantage is described as 'ecological niche inheritance,' leveraging its structural ties within the industrial ecosystem, making it hard for independent companies to replicate.

QWhat are the major practical barriers for companies seeking NVIDIA's AIDC energy storage certification?

AMajor barriers include: 1) High technical requirements for PCS control and integration. 2) The need to submit 12 months of PCS delivery history and a credible 24-month plan for 10x production scale-up. 3) Lengthy (12-24 months) and expensive (potentially millions in fees) multi-standard certification processes covering electrical safety, thermal runaway, and cybersecurity for the entire energy system.

Lecturas Relacionadas

Goldman Sachs In-Depth Report: Who Will Be the Long-Term Winners in China's AI Large Model Industry?

Goldman Sachs Report: China's AI Models at an Inflection Point China's open-source/open-weight large language models (LLMs) have reached performance parity with top global proprietary models, according to a Goldman Sachs report. This is driven by architectural innovations and higher parameter efficiency, allowing Chinese models to achieve comparable capabilities at 2%-10% the parameter size and significantly lower cost. The market is evolving into a two-tiered structure: a high-end segment (e.g., GLM5.2, Qwen3.7 Max) with premium pricing and a low-end, price-sensitive segment for global SMEs and individual users. Key points: * **Cost & Performance:** Innovations like Mixture of Experts (MoE) enable high performance with smaller models. Projects like Meituan's LongCat 2.0, trained on domestic hardware, highlight progress in tech self-sufficiency. * **Open-Source Strategy:** Most Chinese players use open-source/open-weight models for flexibility and ecosystem growth. However, Goldman notes this may underreport actual deployment and revenue. A shift toward "open-weight + community license" models with revenue sharing (e.g., MiniMax) could improve monetization. * **Market Shift & Global Expansion:** Enterprise AI adoption is shifting from "token maximization" to "ROI-first." International expansion, especially in non-US markets, is a major growth driver. Chinese models are increasingly available on global platforms like AWS Bedrock and Microsoft Copilot. * **Competitive Landscape:** Using a framework based on pricing power, cost advantage, and financial strength, Goldman identifies **Zhipu AI and DeepSeek** as the strongest in foundational text models, and **ByteDance** as the leader in multimodal/video generation. The report maintains Buy ratings on MiniMax and Kuaishou. * **Market Growth:** China's AI model API and subscription revenue is projected to grow from an estimated ¥35 billion in 2026 to ¥879 billion by 2030.

marsbitHace 12 min(s)

Goldman Sachs In-Depth Report: Who Will Be the Long-Term Winners in China's AI Large Model Industry?

marsbitHace 12 min(s)

Goldman Sachs Deep Dive Report: Who Will Become the Long-Term Winners in China's AI Large Model Industry?

Goldman Sachs Report: Who Will Be the Long-Term Winners in China's AI Large Model Industry? China's AI large model sector is at a historic inflection point. Goldman Sachs argues that the intelligence of Chinese open-source/open-weight models is approaching top global proprietary models. Rapid adoption by domestic enterprises and global SMEs is creating a data flywheel effect that will further drive model iteration. The evolution is summarized as moving from "DeepSeek's cost-efficiency moment last year to GLM's model-intelligence moment this year." Chinese models achieve near-state-of-the-art performance at significantly lower cost, primarily due to architectural innovations like Mixture of Experts (MoE) and higher parameter efficiency. Models like DeepSeek V4 Pro (1.6T params), GLM5.2 (0.7T), and MiniMax M3 (0.4T) are much smaller than global leaders. Recent advancements in coding capability are attributed to better data curation and RLHF. Landmarks like Meituan's LongCat 2.0, trained fully on domestic AI chips, demonstrate progress in hardware stack independence. The market is forming a "two-tiered structure." The high-end tier (e.g., GLM5.2, Alibaba's Qwen3.7 Max) prices around $1 per million tokens, about 10-25% of US top models, with estimated inference gross margins of 10-20%. The low-end tier (priced as low as $0.06-$0.2 per million tokens) targets price-sensitive global SMEs and individuals. MiniMax derives 60-70% of revenue overseas. Goldman forecasts China's AI model API/subscription revenue to grow from an estimated RMB 35bn in 2026 to RMB 879bn by 2030. Most Chinese players adopt open-source/open-weight strategies for deployment flexibility and community feedback, though this limits monetization as deployments on third-party platforms (e.g., Alibaba Cloud) may not generate direct revenue. A shift towards "open-weight + community license" models with revenue-sharing agreements (like MiniMax's approach) could improve unit economics. International expansion, particularly in non-US markets, is the key growth driver. The global enterprise AI paradigm is shifting from "token maximization" to "ROI prioritization." Chinese models are already hosted on major global platforms like AWS Bedrock and are under consideration for integration into Microsoft Copilot. Using a competitive framework based on pricing power, cost advantage, and financial strength, Goldman identifies the strongest players: In foundational text models, Zhipu AI (initiated coverage) and DeepSeek lead. In multimodal/video generation, ByteDance's Seed is the frontrunner, with Kuaishou's Kling and MiniMax's Hailuo also well-positioned. Goldman maintains a Buy rating on MiniMax, citing its attractive valuation.

链捕手Hace 16 min(s)

Goldman Sachs Deep Dive Report: Who Will Become the Long-Term Winners in China's AI Large Model Industry?

链捕手Hace 16 min(s)

Is Ethereum Truly a "World Computer"?

Title: Is Ethereum Really a "World Computer"? Ethereum, envisioned as a "world computer" by its founder Vitalik Buterin, aims to be a decentralized platform for global applications. However, a recent analysis by Four Pillars raises questions about whether it is more accurately a "Western computer," based on the geographical distribution of its validators. Currently, the United States dominates with 38.19% of all validators, followed by Germany at 13.04%. Combined, these two countries account for over half of the network. In contrast, Asian representation is minimal, with Singapore holding only 3.15%. The concentration is partly due to affordable cloud hosting services like Hetzner and OVH in Europe and North America, as well as the prevalence of residential validators in the U.S., where individuals run nodes via home internet connections. When examining professionally operated validators, the distribution becomes more balanced. The U.S. share drops to 25.81%, while Asian countries like Singapore (7.28%), Hong Kong (6.44%), Japan (6.38%), and South Korea (4.59%) collectively approach the U.S. level. This shift reflects strategic deployments by institutions to meet regulatory requirements and reduce latency for local users. However, regions like South America, the Middle East, and Africa remain underrepresented. Ethereum's peer-to-peer network mechanisms, such as gossipsub, disadvantage areas with low node density, creating a feedback loop where delayed message propagation reduces validator performance and rewards. This imbalance challenges Ethereum's promises of censorship resistance and global accessibility. Despite these issues, opportunities exist for growth in underrepresented regions. As demand for localized staking infrastructure rises, early entrants in areas like the Middle East could establish dominant positions by offering compliant, low-latency solutions. The evolving validator landscape highlights both the structural challenges and the potential for Ethereum to move closer to its "world computer" ideal.

Foresight NewsHace 2 hora(s)

Is Ethereum Truly a "World Computer"?

Foresight NewsHace 2 hora(s)

Trading

Spot

Artículos destacados

Cómo comprar CHIP

¡Bienvenido a HTX.com! Hemos hecho que comprar USD.AI (CHIP) sea simple y conveniente. Sigue nuestra guía paso a paso para iniciar tu viaje de criptos.Paso 1: crea tu cuenta HTXUtiliza tu correo electrónico o número de teléfono para registrarte y obtener una cuenta gratuita en HTX. Experimenta un proceso de registro sin complicaciones y desbloquea todas las funciones.Obtener mi cuentaPaso 2: ve a Comprar cripto y elige tu método de pagoTarjeta de crédito/débito: usa tu Visa o Mastercard para comprar USD.AI (CHIP) al instante.Saldo: utiliza fondos del saldo de tu cuenta HTX para tradear sin problemas.Terceros: hemos agregado métodos de pago populares como Google Pay y Apple Pay para mejorar la comodidad.P2P: tradear directamente con otros usuarios en HTX.Over-the-Counter (OTC): ofrecemos servicios personalizados y tipos de cambio competitivos para los traders.Paso 3: guarda tu USD.AI (CHIP)Después de comprar tu USD.AI (CHIP), guárdalo en tu cuenta HTX. Alternativamente, puedes enviarlo a otro lugar mediante transferencia blockchain o utilizarlo para tradear otras criptomonedas.Paso 4: tradear USD.AI (CHIP)Tradear fácilmente con USD.AI (CHIP) en HTX's mercado spot. Simplemente accede a tu cuenta, selecciona tu par de trading, ejecuta tus trades y monitorea en tiempo real. Ofrecemos una experiencia fácil de usar tanto para principiantes como para traders experimentados.

436 Vistas totalesPublicado en 2026.04.21Actualizado en 2026.06.02

Cómo comprar CHIP

Discusiones

Bienvenido a la comunidad de HTX. Aquí puedes mantenerte informado sobre los últimos desarrollos de la plataforma y acceder a análisis profesionales del mercado. A continuación se presentan las opiniones de los usuarios sobre el precio de CHIP (CHIP).

活动图片