blog.demetriusnew.com

Tag: DataCenters

  • AI Data Centers Are Becoming Large Enough to Require New Reliability Rules

    The growth of large computational loads is changing the operating relationship between data centers and the electric grid. NERC’s new standards point toward a more formal governance model—one that connects facility operations, engineering, and executive oversight to bulk-power-system reliability.

    A data center can be highly resilient inside its fence and still create risks at the grid interface.

    That distinction matters as AI facilities, data centers, and other computational loads grow in scale and connect to the bulk power system. Their electrical demand can change quickly, and their response during a disturbance can affect more than the facility itself. Grid operators need timely information about the load, credible models of its behavior, and clear communication with the people who operate it.

    NERC’s proposed computational-load reliability standards are an effort to establish that foundation.

    From stakeholder ballot to regulatory review

    On October 15, NERC announced that the standards and their implementation plans had passed their final stakeholder ballot. On October 26, NERC’s Board of Trustees adopted the standards. NERC has said it will submit them to the Federal Energy Regulatory Commission for approval in the United States. Until FERC approves them and the applicable implementation period arrives, they are not enforceable U.S. reliability standards. (nerc.com, nerc.com)

    The proposed registry criteria would apply to Computational Load Owners and Computational Load Operators at sites with total connected load of at least 50 MW, supplied through electrical equipment connected at 100 kV or higher. This is a proposed registration threshold—not a claim that every data center or large electricity customer is covered. The related registry criteria also have a separate approval path. (publicpower.org, nerc.com)

    What the standards address

    NERC’s first phase centers on basic capabilities needed to integrate large computational loads reliably. The proposed standards address how loads are interconnected and represented in system models; how facility operators communicate with relevant grid entities; and how protection coordination and disturbance monitoring support reliable operation. The project materials identify three foundational standards—CLO-001-1, CLO-002-1, and CLO-003-1—alongside related revisions to interconnection requirements and studies. (nerc.com)

    These requirements matter because grid reliability depends on a shared operational picture. If the load changes unexpectedly, operators need to understand what happened, how the facility responded, and whether similar behavior could recur. Reliable models, clear communications, coordinated protection, and usable disturbance records help make that possible.

    The governance question for executives

    The standards create a practical governance challenge: responsibility may be split among the entity that owns the site, the entity that operates it, the data-center tenant, the utility, and other grid participants. NERC’s registry revisions establish distinct Computational Load Owner and Computational Load Operator roles, underscoring the need to identify who is accountable for each operational duty. The proposed registry criteria remain subject to FERC approval. (nerc.com)

    Executives should begin with a clear map of ownership and accountability:

    • Assign a responsible executive. Give one leader authority to coordinate energy, engineering, operations, compliance, and external grid relationships.
    • Clarify owner and operator responsibilities. Document who maintains models and protection settings, who communicates with grid entities, who responds to disturbances, and who preserves records.
    • Assess operational readiness. Review procedures, contact paths, escalation authority, and after-hours coverage. A contact list that works only during business hours is not an operating plan.
    • Review data and model quality. Confirm that technical information shared with grid operators reflects current facility design and actual operating behavior.
    • Connect contracts to operating reality. Agreements with tenants, operators, utilities, and service providers should support timely information sharing and coordinated action.
    • Prepare evidence of control. Keep records that show how the facility manages communications, models, protection coordination, and disturbance response.

    This work is useful before a rule becomes enforceable. The goal is not to guess at final compliance obligations; it is to identify gaps early enough to resolve unclear responsibilities, missing data, and fragile operating procedures.

    A first phase, not the finish line

    NERC describes the adopted standards as the first phase of its broader large-load action plan. Additional standards work is underway around computational-load performance, planning and reliability studies, event reporting, and model quality. (nerc.com)

    That points to a wider shift in how data-center readiness will be judged. Power availability, backup generation, and facility uptime remain essential. But large computational facilities also need to show that their grid-facing behavior is understood, communicated, and governed.

    For boards and executive teams, the strategic question is straightforward: Can the organization explain who is responsible for the facility’s effects on the grid, how those responsibilities are carried out, and what evidence shows the controls work?

    As computational loads grow, reliability will depend on that answer being clear before the next disturbance—not reconstructed afterward.

  • Time to Power: Why Grid Delays Can Become Contract Risk for AI Infrastructure

    AI infrastructure projects are often described through large, attention-grabbing numbers: megawatts, GPUs, investment totals and planned opening dates. But those figures only tell part of the story. For a customer planning cloud or AI services, capacity matters when it is ready to use.

    The planned Nscale site in Loughton, Essex, offers a concrete example of how the schedule can break down.

    Reuters reported that Microsoft withdrew from its agreement for capacity at the site after the expected grid connection slipped into the 2030s. Nscale had planned a facility with 50 MW of AI capacity, scalable to 90 MW, and approximately 23,040 Nvidia GB300 GPUs for Microsoft Azure. The original announcement anticipated GPU delivery beginning in early 2027. 

    Microsoft said its decision related to one site and reaffirmed its wider partnership with Nscale. Nscale said it remained committed to Loughton and was exploring alternative power sources and other UK expansion sites. The public reporting does not disclose the contract’s terms, the exact termination provision or the financial consequences for either party. 

    A project schedule depends on more than construction

    Data centre readiness depends on several linked pieces: the building, electrical equipment, cooling, network connectivity, hardware and sufficient power. If one critical dependency arrives years later than planned, the rest of the investment may not produce its expected value on schedule.

    That creates commercial consequences. A late power connection can postpone the point when a customer can use capacity, delay revenue, extend the time capital is tied up and create uncertainty about whether demand will still be there when the site is ready. These are general project risks; the publicly available details do not establish which of them applied to this specific agreement.

    Why this can become a contract issue

    Long-term infrastructure agreements have to allocate the risks of delay. Executives and project teams should understand how the contract treats power availability, delivery milestones, delay thresholds, customer remedies and termination rights.

    The questions are practical:

    • Is the power date confirmed, estimated or dependent on future grid work?
    • What happens if service begins later than the customer’s required date?
    • Is there a defined long-stop date after which a party can exit?
    • Can capacity be moved to another site or supplied from another source?
    • Who bears added costs while the project waits?

    These are questions to raise during commercial and legal review—not assumptions about the terms of Microsoft’s agreement.

    The executive takeaway

    A planned megawatt figure doesn’t guarantee usable capacity on the date a business needs it. Time to power affects whether a project can support customer commitments and whether its economics still work when the facility is ready.

    The better diligence question is not only, “How much AI capacity will this project deliver?” It is also, “When will power be available, how certain is that date, and what happens contractually if it moves?”

    In AI infrastructure, the power schedule belongs in the commercial plan from the start.

  • The Next AI Infrastructure Bottleneck May Be Electrical Cable

    AI infrastructure is often discussed through its most visible assets: advanced chips, enormous data centers and new sources of power. But those assets only become a functioning system when the supporting electrical infrastructure is available and connected.

    Electrical cable is one of those less visible dependencies—and it may be harder to scale quickly than project announcements suggest.

    On October 22, Reuters reported that Nexans CFO Vincent Piquet said cable production capacity is emerging as a bottleneck for AI-driven data-center construction. He described global capacity as tight and said that establishing a new cross-linked polyethylene (XLPE) cable production line, known in the industry as a CV line, takes around three years. 

    That estimate needs to be read precisely. It concerns the time required to establish a new production line. It does not mean that every cable order has a three-year delivery time, nor does it establish that all cable categories are constrained to the same degree.

    Why cable matters to data-center construction

    A data center is not simply a building filled with servers. It needs electrical systems that bring power into the site and distribute it through the facility. Reuters reported that data centers require large amounts of low- and medium-voltage cable for those connections, with larger facilities consuming proportionately more medium-voltage cable. 

    That creates a timing challenge. Demand for a project’s power infrastructure can rise faster than manufacturers can add capacity. Even when a developer has secured land, financing or computing equipment, the project still depends on the availability of components needed to connect and distribute electricity.

    The mismatch is between project urgency and industrial lead time. Data-center developers may want to move quickly, while manufacturing expansion requires investment, equipment, skilled labor and time to bring new lines into service.

    Procurement is already adapting

    The Reuters report also described a change in procurement behavior: French distribution network operator Enedis has moved from annual agreements to contracts lasting six or seven years to secure cable supply. That is one example of a buyer seeking more certainty through longer commitments; it should not be generalized to every market or customer. 

    Still, it points to a wider planning question. When essential components are scarce or slow to expand, procurement becomes part of infrastructure strategy. Buyers may need earlier demand forecasts, longer supplier relationships and clearer visibility into which projects will actually proceed.

    Where the opportunity may be

    The opportunity is broader than cable manufacturing alone. It may include:

    • Manufacturers able to expand the specific cable capacity the market needs.
    • Suppliers of specialized materials, production equipment and related components.
    • Procurement teams that secure supply early and coordinate across project schedules.
    • Developers and utilities that identify long-lead electrical components before construction is underway.

    These are possible areas of opportunity, not guarantees of demand or financial returns. The specific product, market, customer commitments and ability to deliver all matter.

    Nexans’ own October 22 financial release provides a concrete company response: it said it is investing to expand medium-voltage production capacity in Europe and the United States, with data centers among the markets it aims to serve. That confirms Nexans is adding capacity in that area; it does not show that the broader supply constraint has been resolved. 

    The bigger lesson for AI scaling

    The cable story fits a recurring pattern in infrastructure development: the component that limits progress may not be the one receiving the most attention.

    AI capacity depends on a sequence of linked systems—generation, grid connections, substations, electrical distribution, cooling, buildings and computing equipment. A delay in one essential layer can affect the schedule of the whole project.

    For executives, the practical question is not only how much capacity has been announced. It is whether the critical components can be manufactured, delivered and installed on the timeline the project assumes.

    A data center can be planned in megawatts. It still has to be built one connection at a time.

  • AI Infrastructure Debt Is Becoming an Executive Risk Indicator

    The market for AI infrastructure is sending a signal executives should watch: the cost of financing the buildout can change faster than the infrastructure can produce revenue.

    On October 16, Reuters reported that CoreWeave bonds issued in June at a 9.675% yield had repriced to roughly 15%. The yield briefly moved above 16% during the week. That shift reflects investors demanding much more compensation to hold the company’s debt amid concerns about leverage, borrowing costs, execution risk, and the time required for large infrastructure investments to generate returns. reuters.com

    For executives, the point is not that one bond price predicts the future of AI. It is that financing conditions are becoming a useful indicator of which projects and business models the market considers more exposed. A data center can be strategically important and still face pressure if its debt costs rise before customer revenue catches up.

    That timing matters. AI infrastructure requires substantial investment in chips, power, cooling, buildings, and connectivity. Those assets may take years to plan, build, connect, and bring into productive use. During that period, a company still has to fund construction, service debt, and manage the possibility that equipment or market conditions change.

    A higher market yield does not automatically mean the company will pay that rate on all of its debt. But it can make future borrowing more expensive, complicate refinancing, and influence how investors assess new projects. As credit tightens, lenders may favor projects closer to completion, with clearer paths to service and revenue. Reuters reported that investors were showing preference for data center financings with fewer hurdles before entering operation. reuters.com

    This is a financing signal, not proof that AI demand is collapsing. Demand can remain strong while the economics of building capacity become harder to underwrite. The distinction matters: demand answers whether customers want compute; credit markets help answer whether the infrastructure can be financed at an acceptable cost and on a viable timeline.

    Executives tracking the market should watch several indicators together: borrowing costs, debt maturities, interest expense, construction and grid-connection timelines, customer commitments, and the pace at which new capacity becomes revenue-producing. A rising yield on one issuer is a data point. A pattern across financing costs, project delays, and customer economics would be a stronger signal.

    The executive question is not simply, “Is AI demand growing?” It is: Can this infrastructure be financed, delivered, and monetized before its costs or assumptions change?

    AI infrastructure debt is becoming an executive risk indicator because it makes the distance between planned capacity and productive capacity visible to the market.

  • The Next AI Power Plant May Not Look Like a Traditional Power Plant

    AI infrastructure is changing the conversation about electricity supply. Data centers need large, dependable amounts of power, while new generation and transmission can take years to plan and build. As a result, utilities and technology companies are considering a wider set of generation options.

    American Electric Power’s planned fuel-cell project near Cheyenne, Wyoming, is one example. Announced on October 14, the project is expected to provide 702 megawatts of generation for Google under a 20-year power purchase agreement. AEP estimates the investment at approximately $4 billion. Construction is planned to start in 2026, with the facility expected to be fully operational by the end of 2028. The project company, Cheyenne Fuel Cell LLC, is approximately 92% owned and controlled by AEP, with the remaining ownership held by funds advised by Basalt Infrastructure Partners. 

    Why fuel cells enter the AI power conversation

    Fuel cells generate electricity through an electrochemical process. Unlike a combustion-based generator, they do not burn fuel to produce electricity. Bloom Energy’s solid oxide systems can operate on natural gas, biogas, or hydrogen. This gives developers several possible fuel pathways, though the project’s actual fuel choice and supply arrangements matter to its cost and emissions. 

    The technology can therefore broaden the generation toolkit. A large customer may be able to contract for a dedicated source of on-demand electricity without relying solely on new conventional generation or waiting for grid upgrades to serve every increment of demand. That does not mean fuel cells replace the grid. It means they may become one part of a portfolio that includes grid supply, renewables, storage, conventional generation, and other emerging technologies.

    The scale of this project is also significant. At 702 MW, it is not a small backup installation. It is a large generation development with a substantial capital commitment and a long-term customer relationship. The structure links an energy supplier, a technology provider, an infrastructure investor, and a major electricity customer.

    The contract creates visibility, not certainty

    A 20-year power purchase agreement provides a long-term commercial framework. It signals that Google expects a sustained electricity need in the Cheyenne area and that AEP sees an opportunity to invest in dedicated generation for a large customer. Reuters also reported that Google has a separate arrangement with Black Hills Corp. for generation supporting its proposed data center in the area. 

    However, a long-term agreement does not by itself establish the final price of electricity, the project’s full risk allocation, or its performance over time. The economics will depend on details such as fuel procurement, financing, maintenance, operating costs, construction execution, and the terms that govern availability and performance. Those details are not all provided in the public announcement.

    Emissions claims need to be measured in operation

    AEP says the technology generates electricity without combustion of natural gas and that the project includes plans to integrate carbon-capture technology, with the goal of delivering near-zero-emissions power. Those statements describe the project’s design intent. They are not a published, measured operating emissions result. 

    Fuel-cell generation should not be described as inherently zero-carbon. Bloom notes that systems running on natural gas produce carbon emissions, while fuel choice affects the emissions outcome. The project’s eventual impact will depend on what fuel it uses, how that fuel is sourced, and whether carbon capture is installed and performs as expected. 

    For executives and policymakers, the useful questions are concrete:

    • What fuel will the facility use, and how secure is the supply?
    • What will the delivered power cost over the contract term?
    • What availability and reliability levels are guaranteed?
    • What emissions will be measured at the facility, including any upstream fuel impacts?
    • What carbon-capture system is planned, and how will its performance be verified?
    • What happens if construction, fuel delivery, or equipment performance falls behind plan?

    The broader signal: diversify generation, then verify performance

    The AEP–Google project illustrates a broader shift. AI-related electricity demand is encouraging new commercial arrangements and more varied generation technologies. The potential value of fuel cells is not that they solve every power constraint. It is that they add another possible source of dependable generation that can be evaluated alongside other options.

    The decision standard should remain consistent across technologies: compare delivered cost, dependable output, fuel risk, construction schedule, emissions, and operational performance. A promising configuration becomes useful infrastructure only when it can demonstrate those outcomes.

    The power plant of the AI era may look different from the traditional model. Its business case still has to pass familiar tests.

    How much dependable power will it deliver, at what cost, with what fuel and emissions profile—and can the project prove those results over time?

  • Why a Gigawatt of Electricity Doesn’t Mean a Gigawatt of AI Compute

    A data center’s power capacity tells us how much electricity it can draw. It does not tell us how much useful AI work the facility will produce.

    That difference matters as companies compete for power, build larger campuses, and commit billions of dollars to AI infrastructure. A megawatt secured at the grid connection is only the beginning of the measurement.

    Power has to travel through the system

    Electricity entering a data center must be converted and distributed before it reaches computing equipment. The facility also needs cooling, monitoring, backup systems, and other infrastructure to operate reliably.

    Then there is the equipment itself. A GPU can be installed and powered without running at full utilization. A poorly configured rack, a failed component, or a system restart can reduce productive output even while the facility continues to draw power.

    The result is a gap between electricity consumed and useful computation delivered.

    What the reported 60–70% estimate tells us

    On October 14, TechRadar reported remarks from NVIDIA representatives at a Vertiv event suggesting that roughly 60–70% of electricity drawn by some AI data centers supports AI processing. The report described the remaining share as associated with cooling, power distribution and conversion, equipment configuration, failures, and restarts.

    That figure is an industry estimate reported from executive remarks. It is not an independently audited average for all data centers. Facility design, workload, equipment, climate, operating practices, and the way “AI processing” is measured can all affect the result.

    The estimate is still useful as a prompt: capacity figures alone do not reveal how efficiently a facility turns electricity into productive work.

    “Non-compute” power is not automatically waste

    Cooling and backup systems are necessary to keep high-density computing equipment within operating limits and protect service continuity. Power conversion and distribution are also part of delivering electricity safely to the IT load.

    The goal is not to eliminate every watt that does not go directly to a processor. The goal is to understand the full system, reduce avoidable losses, and make sure the infrastructure investment supports dependable output.

    A facility designed for peak demand and contingencies may also have capacity that is not used continuously. That may be appropriate for resilience, but it raises a management question: how much of the available infrastructure is producing useful work under typical operating conditions?

    Measure the output, not just the input

    “Gigawatt of AI compute” is shorthand. A gigawatt measures electrical power; compute output is measured through completed work and system performance.

    For executives, that means asking for more than a facility’s contracted power or nameplate capacity. Useful measures could include:

    • Electricity delivered to the facility and to IT equipment
    • AI workload throughput, such as completed jobs or tokens generated
    • Accelerator utilization and periods of idle or degraded operation
    • Cooling and power-system overhead
    • Availability, restart frequency, and time spent operating below target

    These measures need consistent definitions and a clear time period. A peak result, a design target, and an observed operating average are different kinds of evidence.

    The strategic implication

    When power is scarce, improving the amount of useful computation produced from existing electricity can be as important as securing more supply.

    That makes infrastructure efficiency an operating and capital-allocation issue. Better power delivery, cooling, equipment configuration, workload scheduling, and maintenance may help increase productive output without assuming that every problem can be solved by adding another megawatt.

    The reported 60–70% estimate should be treated as a signal to investigate, not a benchmark to apply blindly. Each operator needs facility-level evidence to understand where power goes and what work it enables.

    The executive question is therefore not only, “How much power can this data center draw?”

    It is: How much useful, reliable AI work does each megawatt deliver—and how do we know?

  • The Grid Can Track AI Demand. But Who Has Authority to Control It?

    FERC’s PJM decision separates load visibility from curtailment authority

    The growth of AI infrastructure is making electricity demand harder to forecast and plan around. Large data centers can require hundreds of megawatts, ramp up over time, and depend on power arrangements that affect regional reliability planning.

    On October 9, 2026, the Federal Energy Regulatory Commission (FERC) took a step toward improving visibility into that demand. It approved PJM Interconnection’s definition of a “Large Load” and its proposed Large Load Registry. At the same time, FERC rejected much of PJM’s broader proposal for providing interim service to new large loads, including provisions related to curtailment. (elibrary.ferc.gov, utilitydive.com)

    The decision highlights an important regulatory distinction: collecting information about large electricity users is different from deciding when their service can be interrupted.

    What the registry is meant to show

    PJM’s approved definition covers end-use customer load with a cumulative peak of at least 50 megawatts at a single electrical site. The registry is designed to collect information such as a facility’s location, ramp schedule and capacity supply. That information can help PJM understand how large loads may affect forecasts and reliability planning. (elibrary.ferc.gov, utilitydive.com)

    A registry improves visibility. It can help answer questions such as:

    • Where is major new demand expected to connect?
    • How quickly is that demand expected to ramp up?
    • What capacity supply is associated with the load?

    Those are important planning questions. But data collection does not itself create an operational curtailment rule or settle which regulator has authority over a retail customer’s service.

    Why FERC rejected the broader proposal

    PJM’s filing also proposed an Interim Resource Adequacy Service framework for certain new large loads. Under parts of that proposal, eligible loads that did not bring their own new capacity could face load-reduction requests as the system approached an emergency.

    FERC rejected that framework. The Commission said several provisions appeared to place obligations directly on large loads that are, for the most part, retail customers; to dictate which retail customers would be curtailed; and to establish a retail compensation rate. It also found that PJM had not adequately shown that the proposed treatment of those loads would avoid undue discrimination. (utilitydive.com)

    This is where jurisdiction matters. FERC oversees interstate transmission and wholesale market rules. States have a central role in regulating retail electricity service, including determining which retail customers may be curtailed when the system is stressed. FERC’s order treated the issue as a shared regulatory challenge, not as a simple question of whether grid operators can see or manage demand. (utilitydive.com)

    What the decision does—and does not—mean

    The order does not mean large data centers are exempt from every form of curtailment, or that emergency load reduction is off the table. It addresses the specific tariff framework PJM proposed in this filing. FERC also gave PJM direction on issues it would need to address if it revises and resubmits parts of the proposal. (utilitydive.com, headwaterswire.com)

    The distinction is between three separate layers:

    1. Visibility: identifying large loads and gathering information about them.
    2. Grid operations: maintaining system reliability as supply and demand change.
    3. Decision rights: determining which customers can be asked or required to reduce demand, under what rules, and with what protections.

    The registry advances the first layer. The order shows that the third requires careful coordination across federal and state authority, as well as rules that can be defended as fair.

    The governance question behind AI power demand

    For AI infrastructure, the practical issue is not only how much electricity a facility may use. It is also whether the grid and regulators have clear, workable rules for what happens when demand exceeds available supply.

    That requires a complete chain: reliable information about large loads, transparent criteria for reliability actions, clear responsibility for issuing those actions, and consistent treatment across customer classes. Without that chain, the grid may know where demand is growing while decision-makers remain uncertain about who can require it to change.

    The registry can help the grid see the load. It cannot, by itself, determine who has authority to control it.

    Source: FERC, Docket ER26-3515-000, order issued October 9, 2026.

  • The Next AI Infrastructure Bottleneck May Be Chips Waiting on Other Chips

    AI infrastructure discussions often begin with the most visible asset: the accelerator. How many GPUs or other AI processors can a company buy, and when can they be installed?

    But installed capacity is only part of the equation. Those processors must also exchange data, coordinate work, and receive instructions from the software stack. If those connections cannot keep pace with the workload, adding more processors may not deliver a proportional increase in useful compute.

    That makes networking and integration a growing part of the economics of AI infrastructure.

    Compute depends on more than the processor

    An AI cluster is a system of connected components. Processors move data between memory, storage, and other processors while workloads are divided, coordinated, and recombined. The network affects how quickly those exchanges happen and how well the system handles congestion or failures.

    Two networking jobs matter here. Scale-up networking connects accelerators within a pod or rack. Scale-out networking connects systems across racks and throughout a data center. A workload can depend on both.

    When data movement or coordination becomes a constraint, some processors may wait for information or for other parts of the workload to finish. The hardware is present, but its potential output is not being fully realized. That can weaken the return on expensive infrastructure investments and complicate capacity planning.

    What Token Fabric is trying to do

    On October 8, Nvidia-backed Upscale AI introduced Token Fabric, a networking architecture intended to support AI processors from different suppliers. The company describes it as a combination of scale-up and scale-out networking, systems, and shared software for managing operations. Reuters reported that the platform is expected to roll out in stages, with the first component planned for the fourth quarter and the broader platform through 2027. Upscale says general availability is planned for early 2027. (reuters.com, upscale.com)

    The proposal addresses a practical challenge for operators: networking has often been assembled from separate components that customers must integrate and manage. A more unified approach could simplify parts of that work and make it easier to build clusters containing different types of accelerators.

    That is the opportunity. It is not yet proof that the architecture will deliver a specific utilization gain or make mixed-vendor systems simple to operate. Those results will depend on deployment conditions and independent performance evidence.

    Interoperability is useful, but it is not automatic

    The possibility of using multiple accelerator suppliers is attractive for organizations that want more choice in hardware procurement or want to match different processors to different workloads. It could also reduce reliance on a single hardware ecosystem.

    However, networking compatibility is only one part of interoperability. A system still needs suitable software frameworks, drivers, compilers, memory behavior, and workload orchestration. Operators must also understand how performance changes when different processors share a system.

    In other words, “connects different processors” does not necessarily mean every processor can be substituted freely, or that every workload will run equally well across all of them. The relevant test is whether the full system performs reliably and economically for the workloads a customer actually runs.

    The business metric is useful output

    For decision-makers, the central question is not simply how many processors are installed. It is how much useful work the organization gets from the complete system, at what cost, and with what operational burden.

    That means evaluating utilization alongside measures such as workload throughput, latency, energy use, network performance, software compatibility, and the cost of operating the environment. A network upgrade that improves one technical metric may not improve the business outcome if it adds complexity or fails to support the applications that matter.

    The same discipline applies to claims about future capacity. Product roadmaps and announced availability dates are signals about where vendors are investing. They are not substitutes for customer deployments, workload-level results, and independently reviewable performance data.

    The next bottleneck is system-level

    AI infrastructure is moving beyond a simple race to install more chips. The economics increasingly depend on how compute, networking, software, power, and operations work together.

    Token Fabric is one example of vendors trying to address that system-level problem and support more heterogeneous infrastructure. The broader industry question is whether these architectures can make processors easier to connect and keep productively occupied—without creating new integration and management burdens.

    The next generation of AI capacity may be shaped not only by how many chips organizations can obtain, but by how effectively those chips can work together.

  • The Compute Grid: What Happens When Idle GPUs Become Shared Infrastructure?

    AI’s next capacity gain may come from coordinating existing infrastructure more effectively. The harder question is whether shared capacity can be made dependable enough to use.

    AI compute is often described as scarce—and for many organizations, it is. But scarcity can exist alongside underused equipment. Some data centers may have capacity that is idle between workloads, while startups and researchers struggle to secure the right chips at the right time.

    The National Compute Grid initiative, announced October 7, aims to connect capacity across providers through shared scheduling. The idea is to make it easier to find available computing resources and match them with workloads, potentially improving access to infrastructure that already exists. 

    That could add a useful layer to the AI infrastructure market. It does not mean every idle GPU can immediately be combined into a reliable pool—or that shared scheduling makes new data centers unnecessary.

    The opportunity is coordination

    A GPU sitting unused at one facility cannot automatically help a team whose workload is running elsewhere. Capacity is distributed across providers, hardware types, locations, software environments, and commercial agreements.

    A shared scheduling layer could help bridge some of those gaps. In principle, it could show what capacity is available, help match workloads to suitable hardware, and coordinate access across participating providers. The initiative’s organizers have described a network that would pool capacity from labs, cloud providers, and different chip systems. 

    If it works, the value may be more than simply “more GPUs.” Better coordination could make capacity easier to discover, allocate, and use—especially for organizations that cannot secure large, long-term compute contracts.

    “Available” does not always mean usable

    A headline capacity figure can describe resources that are connected, planned, or otherwise in sight. That is not the same as compute that a customer can access continuously, at a known price, with a guaranteed performance level.

    The initiative’s reported figure of roughly 760 megawatts, for example, was described as capacity “connected or in sight,” alongside a goal of reaching 2 gigawatts by 2030. Those are organizer-reported figures, not proof that all of that capacity is currently available to run customer workloads. 

    The initiative’s current website describes its capacity figures as dated planning estimates and says access is limited to public-sector partners during a preview. That is a useful reminder to distinguish a developing network and its stated plans from a mature, broadly available service. 

    The same care applies to utilization statistics. The consortium’s paper reportedly estimates that independent, single-tenant data centers average less than 15% net computing utilization. That figure should not be read as meaning that 85% of all GPUs are sitting idle and ready to be reassigned. Utilization can be measured in different ways, and unused time does not by itself establish that hardware is compatible, accessible, or suitable for another organization’s workload. 

    Workload fit matters as much as capacity

    A pool of GPUs is not automatically a single large cluster. Large training runs can depend on fast networking between thousands of GPUs. If resources are spread across facilities, data movement, network performance, and coordination can limit what workloads are practical. Different chip architectures can also require changes to software and tooling. 

    Shared compute may therefore be useful first for workloads that can tolerate more flexibility: research jobs, inference, or smaller training runs. Whether a distributed pool can reliably support the most demanding jobs remains an operational question.

    The details to watch are concrete:

    • Availability: Can users get capacity when they need it, and for how long?
    • Performance: Does the workload complete at the expected speed?
    • Continuity: What happens if a provider withdraws capacity or a job is interrupted?
    • Compatibility: Can the workload run across the available chip and software environments?
    • Data movement: How long does it take to move datasets to the compute—and what does that cost?
    • Accountability: Who is responsible when capacity is unavailable or performance falls short?

    These are not minor implementation details. They determine whether shared capacity behaves like a dependable service or simply appears in a capacity forecast.

    A new infrastructure layer, with evidence still to build

    The National Compute Grid points toward a broader market question: can software coordination make existing infrastructure more productive and more accessible?

    If the answer is yes, some organizations may get useful compute without building their own facilities or negotiating large dedicated contracts. That could expand access and improve the use of resources already deployed.

    But the evidence that matters will come from operations: how much capacity is actually available, what workloads it supports, how reliably jobs complete, and whether performance and cost meet users’ expectations.

    The next phase of AI infrastructure may include both new buildings and better coordination of existing ones. A shared scheduler can help connect supply and demand. It cannot make every GPU interchangeable—or turn a planning estimate into dependable compute by itself.

    The measure of success is not how much capacity a grid can name. It is how much useful work users can reliably complete with it.

  • The Electrical Bottleneck Inside the AI Data Center

    AI power discussions often begin outside the data center: Can the grid serve the load? Is generation available? How long will interconnection take? Can transmission capacity reach the site?

    Those questions remain central. But they aren’t the only ones.

    After electricity reaches a campus, it must still pass through the facility’s electrical infrastructure and arrive at the equipment that performs the computing. That pathway has physical limits. Conductors, switchgear and other equipment take up space, require installation labor and must operate reliably under demanding conditions. As data centers add more power-dense computing, the electrical system inside the facility becomes part of the scaling challenge.

    VEIR’s October 8 financing announcement is a signal that investors see an opportunity in this layer. The company announced a $110 million Series C, bringing its reported total capital raised to $225 million. It says the funding will support manufacturing, product development and preparation for commercial deployments. VEIR also reports a 3-megawatt demonstration and says it is working with customers and industry partners on reference designs for AI facilities. 

    The approach centers on superconducting power-delivery systems. In VEIR’s description, the system includes more than the superconducting cable: it also requires cryogenic cooling, controls, terminations and monitoring. The goal is to move more power through a smaller physical footprint than conventional power-delivery infrastructure. 

    That could matter where space and power density are becoming linked. A smaller footprint could give designers more flexibility inside a facility or campus. But a compact cable alone doesn’t establish an operational advantage. The complete installation has to work as a system, and its economics must account for cooling, controls, installation, maintenance and the consequences of equipment failure.

    The technology also addresses a specific point in the power chain. It may improve how electricity is distributed within a data center. It does not supply the electricity, secure an interconnection or expand the grid. Those are separate constraints, and they remain part of the larger project.

    That distinction is useful because “AI needs more power” can make every proposed solution sound interchangeable. They aren’t. Some investments add generation. Some expand transmission. Others improve the way power is delivered inside a facility. A project may need progress at several of these layers at once.

    The commercial evidence is still developing. VEIR’s reported 3-megawatt demonstration is a technical milestone, while the company’s announcement describes initial commercial deployments as something it is preparing for. Reuters reported earlier this year that superconducting cables were not yet being used in data centers and that cost and manufacturing constraints had limited adoption. 

    For operators and infrastructure investors, the next questions are practical:

    • How reliably does the full system operate under real data-center loads?
    • What does it cost to install and maintain compared with conventional alternatives?
    • How does the cooling and control equipment affect operating complexity?
    • Can the system be manufactured, installed and serviced repeatedly at commercial scale?

    The $110 million round indicates investor confidence in the opportunity and provides capital to pursue commercialization. The test of the thesis will be what commercial deployments show about reliability, economics and repeatability.

    AI infrastructure can be constrained at more than one point between the grid and the server. Power availability gets attention because it is visible and difficult to solve. Power delivery inside the facility may be less visible, but it can also shape how densely and efficiently a campus grows.

    The useful question is not just how much electricity a data center can secure. It is how dependably and economically that electricity can be delivered to the compute.

    Source note: VEIR’s financing and technical milestones are company-reported; treat commercial performance claims as unproven until supported by operating data.