For the first three years of the generative AI wave, the output of a model was a string. You called an API, you got tokens back, you rendered them in a chat window or wrote them to a row in Postgres. The economics of that pipeline were dominated by inference cost. Storage was a rounding error.

That era is over. The output of a modern generative pipeline is an asset: a 4K video clip, a stem-separated audio track, a 50-megapixel product render, a 3D mesh with PBR textures. Generative AI has gone from text-centric to asset-centric, and the architectural center of gravity has moved with it. The teams building durable advantages in generative media right now are the ones treating their storage layer as a pipeline component rather than a destination.

This is a good problem. It is the problem you get when your pipeline works.

Asset-centric changes the shape of the system

Text pipelines are stateless in practice. A prompt goes in, a response comes out and the interesting state lives in a database. You can rebuild almost any artifact by re-running the call.

Media pipelines are not like that. Every stage produces a large, opaque binary that the next stage consumes. A single finished deliverable might traverse a dozen of them: prompt expansion, base generation, upscale, frame interpolation, color pass, audio generation, mix, mux, transcode to delivery formats, thumbnail extraction. Each stage writes an intermediate. Each intermediate is expensive enough to regenerate that you keep it.

The result is a system where the objects are the state. Your object store stops being a place you put things when you are finished and becomes the substrate the pipeline runs on.

Adoption is past experimentation, and the volume is in production

Advertising has the clearest numbers. IAB’s 2026 Digital Video Ad Spend and Strategy Report finds that nearly two in three digital video buyers now use generative AI for creative, up from half in 2025. A third of their ad assets are expected to involve generative AI this year, up from a quarter in 2025, with buyers projecting 43 percent by 2027. That is happening inside a U.S. digital video ad market IAB projects will pass $80 billion in 2026, growing 11 percent year over year, nearly 20 percent faster than the total ad market.

The interesting detail for architects is what the creative is used for. IAB’s prior-year data showed buyers reaching for generative AI specifically to produce audience-specific versions of an ad, visual style variations and contextually adapted cuts. That is not one asset per campaign. That is a matrix.

Games are the instructive counterexample. GDC’s 2026 State of the Game Industry puts generative AI use at 36 percent of industry professionals and 30 percent at game studios specifically, but the usage breakdown is dominated by language models rather than media generation: research and brainstorming at 81 percent, code assistance and routine writing at 47 percent each, prototyping at 35 percent. The most-used tools are ChatGPT, Gemini and Copilot. Sentiment is sharply negative, with 52 percent saying generative AI is having a negative impact on the industry, rising to 64 percent among visual and technical artists. Asset-centric pipelines have not landed in games the way they have in advertising, and the constraint is as much workforce and provenance as it is tooling.

Elsewhere, the pattern holds even where the survey data is thinner: e-commerce teams generating on-model imagery per SKU per segment, localization pipelines producing dubbed and lip-synced variants per market, previsualization work that used to require an art department.

What the adopting categories have in common is that none of them produce one asset per request. They produce a set. The pipeline is judged on how many viable options it surfaces, which means a better pipeline is, definitionally, a pipeline that writes more bytes.

Reactive architecture, because the model layer will not hold still

The model landscape resets on a cadence measured in weeks. A new video model ships with better temporal coherence. A new audio model handles multilingual prosody properly. A new image model finally gets text rendering right. If your pipeline requires an engineering sprint to adopt a new model, you are structurally behind teams whose pipelines do not.

Reactive architecture is the answer, and it means two specific things.

1. Model-agnostic stages. Each stage of the pipeline should express a contract in terms of inputs and outputs, not in terms of a vendor. A generation stage takes a prompt and conditioning assets and produces a video at a declared resolution and duration. Which model backs it is configuration. Swapping providers should be a config change and an eval run, not a refactor.
2. Event-driven orchestration. Polling-based orchestration couples your stages to a scheduler and makes each new stage a change to the control plane. Event-driven orchestration inverts it: a stage completes, it writes its output, the write itself is the signal that the next stage should start. Adding a stage means subscribing to an event, not modifying a DAG definition that six other teams depend on.

This is where storage stops being passive. Object storage that emits events on write lets your bucket act as the message bus for the pipeline. B2 Event Notifications send a signed HTTP POST to a webhook endpoint when objects are created, updated or deleted, with rules scoped per bucket and filterable by prefix. That prefix filter is the part that matters architecturally: if your bucket is organized by stage, a rule on stage/upscale/ is a subscription to “upscale finished” without any code knowing what upscale is. Custom headers on the notification carry auth tokens or context to the target, so the endpoint can be a queue, a serverless function or a workflow platform rather than a service you had to build.

A completed upscale triggers the color pass. A completed mux triggers the transcode fan-out. A completed transcode triggers the CDN warm and the catalog write. The storage layer sequences the work, which removes an entire class of orchestration glue from your codebase and removes polling latency along with it.

Quality improvements arrive as file size increases

Every generation of media models improves along axes that all cost bytes. Resolution goes up. Frame rate goes up. Duration limits extend. Bit depth and color fidelity improve. Audio moves from mono to multi-channel. Compression artifacts that were acceptable at 720p are not acceptable at 4K, so teams move to higher bitrates and, for anything entering a post pipeline, to intermediate codecs.

The arithmetic is worth doing explicitly. A 10-second clip in a delivery-grade H.264 4K encode at 50 Mbps is roughly 60 MB. The same ten seconds as a ProRes 422 HQ intermediate, which Apple targets at 884 Mbps for 3840×2160 at 30p, is 1.1 GB. That is roughly 18 times the size, and intermediates are exactly what you keep between stages. Now assume your pipeline generates eight candidates per brief because your creative director wants options, and each candidate produces four intermediates before final. That is one brief consuming tens of gigabytes.

Nobody plans for that in a proof of concept. Everybody encounters it in month four of production.

The iteration multiplier

Here is the part that surprises teams: robustness and storage growth are the same curve.

A fragile pipeline produces one output per request because that is all it can manage. A robust pipeline produces candidates, keeps the rejects for training and eval, versions every asset so a creative decision can be reverted, retains intermediates so a late note does not require regenerating from the prompt, and derives proxies, thumbnails and per-platform cuts from every approved master.

Each of those is the correct engineering decision. Together they mean that improving your pipeline increases your storage footprint superlinearly relative to your output volume. If your unit economics assume storage scales with delivered assets, they are wrong. Storage scales with attempts multiplied by stages multiplied by versions multiplied by derivatives.

This is why storage strategy has to be a design input rather than a line item you discover on an invoice. The two things that turn it from a manageable cost into a structural problem are egress pricing and the absence of a lifecycle policy. Egress hurts most in the exact architecture described above because a multi-stage pipeline repeatedly reads its own intermediates, and a distribution layer constantly reads masters. When every read carries a metered charge, the pipeline design that produces the best creative output is also the one that produces the worst bill, and teams start making architectural compromises to protect margin. Lifecycle policy hurts by omission: if you never decide what an intermediate is worth after 30 days, you pay to keep all of them forever.

What to put in place now

If generative media is core to what you are building, four decisions determine whether your storage layer accelerates you or constrains you:

1. Choose a storage economic model that does not penalize reads. Understand your egress terms before your architecture depends on them. A pipeline that reads its own outputs at every stage is a read-heavy workload, and pricing that assumes write-once, read-rarely does not fit it. Model the ratio you actually expect: egress as a multiple of stored volume, not as an absolute. That ratio is the number to design against.
2. Make writes trigger work. Use object-level event notifications as the pipeline’s signaling mechanism. This buys you loose coupling, lower latency between stages and the ability to add a stage without touching the orchestrator.
3. Attach metadata at write time. Model version, prompt hash, parent asset, generation parameters, approval state. Metadata written at generation time is nearly free. Reconstructing provenance across a million objects later is not; provenance is what makes your rejected candidates usable as eval data and a training signal.
4. Define lifecycle policy per artifact class. Masters, approved derivatives, intermediates and rejects have different retention values. Encode that as policy on day one rather than as a cleanup project in year two.

Which points at the useful way to think about the storage layer: in a pipeline where everything else churns, it is the constant. Models turn over every few weeks. Stages get swapped, added and reordered around them. Output volume compounds with every quality improvement. The one layer absorbing all of that without being redesigned is the one holding the assets, so it is worth choosing based on the characteristics that stay true while the rest moves. That is what we built B2 for. Always hot, so no stage waits on a rehydration to read what the last one wrote. No retention minimum or file size floor, so intermediates that were always disposable cost what they used. Egress scales to what you store rather than metered per read, so a pipeline that reads its own output is not penalized for being good at its job. The architecture above it should change every quarter. The storage underneath it should not have to.

The opportunity

The teams that will win in generative media are not the ones with privileged access to a model. Model access is converging toward commodities. The advantage is in the pipeline: how fast you can adopt a better model, how many candidates you can afford to generate, how much history you retain to evaluate and fine-tune against, and how cheaply you can move all of it.

Every one of those is a storage architecture question. Treat the storage layer as an active participant in the workflow and it becomes the thing that lets you iterate faster than your competition. Treat it as a bucket you dump finished files into, and it becomes the ceiling on how good your pipeline is allowed to get.

The assets are the product now. Architect accordingly.

Lead times of nine to 12 or even 18 months. Costs rising by 35%, 45%, even 50% to 200%. More than halfway through 2026, the market for IT infrastructure that’s crucial for enterprise projects, including those involving artificial intelligence, is strapped.

Memory is at the root of the shortages. Memory prices “have risen by 50% to 200%, resulting in PC prices increasing by 35% to 45% and some server prices rising over 125%,” according to Jon Forest, VP analyst at Gartner. Network switches also need memory, albeit in lesser amounts than servers, so they are not immune, with prices and lead times likewise rising dramatically.

Industry experts agree that most of the issues stem from hyperscalers gobbling up memory capacity, which trickles down to servers, storage systems, and networking devices. But while the source of the problem may be new, supply chain disruptions are far from unprecedented.

As a result, industry insiders are not short on advice on how best to deal with the situation, with tips including making better use of what you have, considering options beyond your usual scope, and lots of planning with your vendors and internal finance teams.

State of the problem

Just how bad is the current supply chain problem? “It’s pretty bad,” says Matt Kimball, vice president and principal analyst with Moor Insights & Strategy. Companies accustomed to 30- to 45-day lead times for various infrastructure are now looking at 6, 12, or even 18 months.

“It’s real, and I’m hearing it from companies of all sizes, from the 1000-server to the 10,000-server shops,” Kimball says.

“Memory costs are expected to rise sharply well into 2027 and will reach up to 25% of network hardware expenses by the end of 2027,” according to an email Gartner’s Forest sent to Network World. The figure below shows the timeline Gartner expects for memory prices, and Forest notes that the same timing applies across networking, storage, and compute infrastructure. 

“Enterprise network equipment pricing is projected to increase by over 20% in 2026. This upward trend is anticipated to continue with a further rise of 3% to 5% entering 2027, with no signs of price reduction until the end of 2027.”

But “reduction” will likely look more like “stabilization.”

“That’s something a lot of people don’t like to talk about. But let’s say prices went up 40%, they may come down five,” says Phillip Privett, senior vice president of vendor management with the global distributor and value-added reseller TD SYNNEX. “They’re not going to come down 40%.”

Perhaps worse, compared with past disruptions caused by issues such as fires in chip fabrication factories or the Covid pandemic, Kimball says this one is “durable” because its cause—the AI wave—is more long-lasting and just getting started.

“This AI inference wave we’re hitting is just beginning. It’s going to be longer and bigger than the training wave,” he says. “It’s impacting everything, from AI infrastructure to the traditional stuff that’s standing up your virtualization and cloud infrastructure.”

No vendors seem to be immune, not even the likes of Cisco, which makes its own Cisco Silicon One chips. Or, at least, it designs the chips; they’re actually manufactured by the Taiwan Semiconductor Manufacturing Company (TSMC), the same company that makes many of the other chips that are in such demand. And that’s only one component of many that comprise a switch.

On the other hand, the margins Cisco gets from enterprise sales are far greater than those from hyperscalers because Cisco sells mainly just hardware to hyperscalers, whereas enterprise sales generally include software and services as well. So, Cisco has incentive to keep enterprise customers happy and maintain the 66% margins it reported in Q3, its latest quarter.

Still, Cisco must deal with the same shortages as other vendors.

“I wouldn’t say any company is faring better than others,” says Neil Anderson, vice president and CTO for cloud, infrastructure, and AI solutions at World Wide Technology (WWT). “There may be nuances that some suppliers are employing to balance it to some extent, but I fail to recognize a supplier that’s not having almost the same issue.”

Cloud storage vendor Backblaze is one company that’s facing equipment cost and availability issues. “There are different types of shortages occurring in multiple places, all driven by unusual market demands, really by just a handful of very large buyers,” says James Rowell, senior vice president of operations with Backblaze.

Backblaze is constantly forecasting and monitoring demand triggers, Rowell says. That involves close alignment with the sales team to forecast client needs, as well as paying attention to historical trendlines to predict upcoming demand from new deals and growth with existing clients. But the company also looks for “unnatural market-related triggers” that would cause a spike in utilization.

With hyperscalers buying up vast amounts of capacity, “This is definitely an unnatural phase,” Rowell says. “For about for the last 12 months, I would say there’s been somewhere between a 15% and 30% uptick in costs,” especially in terms of servers and compute disks.

On the positive side, at least for Backblaze, the company is also seeing an uptick in business from an interesting source: AI companies. “We reported in the last earnings period a 70% increase in AI companies using our platform,” says Patrick Thomas, vice president of marketing at Backblaze. “That’s massive.”

On top of that, the company is seeing an uptick in deals from enterprises that can’t get the storage capacity they need or want on-prem. “There’s a general market nervousness where we’ve got potential deals coming our way because those organizations are concerned about being able to do it themselves,” Rowell says.

While some expect new chip fabrication plants currently under construction will ease memory supply constraints, Privett doesn’t buy it. “I don’t see it getting better anytime soon,” he says. “Building a new fab is a two-year process.”

Advice: Start with the basics

Enterprises, then, must play the cards they’re dealt. For Moore Insights’ Kimball, who did stints as an IT exec with the states of Florida and Oregon, that starts with making the most of what you have.

Such a strategy is “shockingly not implemented much” across the companies he sees. “A simple capacity planning exercise can free up a lot of resources.” That includes virtualized servers running at just 20% to 30% utilization as well as extending the life of existing servers. While 15 or 20 years ago it was common to refresh every four years or so, companies can often get six or seven years out of today’s servers.

While such strategies won’t solve your AI compute challenges, they can certainly help support your ongoing operations and free up budget for AI and other modernization projects, he says.

“Sweat your assets,” agrees Privett of TD SYNNEX. “Work them as much as you can, add only what you need, get extensions on your licensing, renewals on your services agreements and things like that. Just sweat it out a little longer.”

If you have budget to spend but can’t get the hardware you’re after, buy something else, says WWT’s Anderson. “Look at things that are not tied to those components, like software projects or SaaS licensing,” he says.

Get friendly with finance teams

Numerous experts recommend regular meetings with your CFO or finance teams to keep them apprised of what you’re up against so the company can plan accordingly.

Gartner’s Forest advises using rolling 12- to 24‑month forecasts and engaging early with suppliers to identify constrained components and SKUs. Committing to quarterly or monthly buys can help you avoid long-term agreements that extend past the rapid increases we’re seeing in 2026, he says.

Also engage with the financing arm of your equipment vendors, some of which are offering financing incentives, Privett says. Compute vendors in particular are offering subsidized financing, deferred payments, and low-cost financing for the first year or so. “Those are huge opportunities to take advantage of,” he says.

By engaging with finance teams, IT groups can conduct budget allocation exercises and try to come up with ways to make the financials work. The last thing you want to do is surprise them with additional budget requests out of the blue.

Kimball recalls his days with the state of Florida, when all budget requests were examined by a technical review working group—which was designed to be hostile.

“I can’t imagine going to them and saying, ‘Oh, did I say that was a million dollars? It’s actually $2 million. I need you to write me a bigger check,’” he says. “I would walk into one of the swamps in Tallahassee and get eaten by the alligators instead of doing that.”

Work with your vendors and VARs

As you put plans together, lean on your vendors for help, including channel partners such as value-added resellers (VAR) and national resellers. “Work with them to map things out and understand what your workloads will look like,” Kimball says.

That’s what Backblaze’s Rowell regularly does with his suppliers. He lays out his forecast for the year, with commitments on what Backblaze will definitely buy, as well as scenarios that account for rapid growth, say, 2x. “And they’ll come back with, ‘Well, okay, no problem,’ or maybe they say we need to put in an allocation right away, or we won’t be able to get what we may need,” he says.

Similarly, he sits down with his CFO regularly to map out predictive models that factor in inflation, price hikes, and the like. The idea is to plan out multiple scenarios, so you don’t get blindsided.

“If you don’t do that, you’ll get caught with your pants down, on the upside-down end of spectrum,” he said – meaning not having the capacity to take advantage of market opportunities.

Acquiring the capacity you need to meet project demand may also mean being flexible in terms of your equipment choices. If you’re a Dell shop but can’t get Dell servers, maybe you go with Lenovo, Kimball says.

“You’ve got to figure out how to use all this silicon and infrastructure in a heterogenous way to serve your needs,” he says. That’s especially true when it comes to AI infrastructure. “If you think you’re going to go with 100% Nvidia for everything from RAG [retrieval augmented generation] to inferencing at the edge, you’re kind of crazy, not because of cost but because of availability.”

Look at alternatives, including AMD and cloud solutions, while staying mindful of how it all plays together. You may not be able to get Nvidia GPUs, but AWS, Azure, and Oracle Cloud have them, Kimball notes.

Be strategic, perhaps by using cloud offerings to handle certain tuning or inference workloads, then bringing them back in-house when appropriate. “Have a better understanding of what absolutely has to be on prem and what can be in the cloud,” he says.

That’s good advice, says Backblaze’s Thomas. When it comes to AI, think about performance tiers and the range of use cases you have. They don’t all need top-tier performance.

“People get wrapped around axle of needing the top end. There’s a lot of flexibility in the edges, innovation in different hardware and software,” Thomas says.

Gartner likewise advises companies to increase configuration flexibility and expand sourcing paths. That may include buying from secondary markets and lease-return programs to preserve continuity with existing infrastructure until the shortages pass, Forest says.

Get started somewhere

Even if you can’t acquire or have to wait for the infrastructure you need, don’t let that keep you from getting started with AI or other modernization projects.

Options include public cloud and neocloud providers, Anderson says. WWT also provides capacity in its own lab so customers can get started with proof-of-concept projects. “Don’t just throw your hands up. We can help you find access to capacity,” Anderson says. “Production-scale AI may be delayed, but don’t let that derail your strategy.”

Colocation providers may likewise be an option, especially if enterprises are struggling to acquire high-end networking equipment. Networking is a key value proposition for colocation providers, in that they have built-in connections to various cloud providers and other ecosystem players.

Equinix, for example, has 280 data centers in 77 metropolitan areas, says Phil Read, senior director, colocation product management for the company. If you have the compute infrastructure, Equinix can help you with the high-end connectivity required both intra- data center and at edge facilities.

It also has partnerships with the likes of Cisco and Nvidia for “ready-to-go AI connectivity,” Read says. That means Equinix offers the right infrastructure to meet the requirements of high-end compute solutions in terms of power density and cooling. Such power densities are significant, requiring 120k VA per rack and up. “There’s plenty of talk about a megawatt rack,” he says.

Power is a significant issue in this entire discussion, Privett says. Older installed computing infrastructure likely consumes far more power than newer systems, which is an argument for upgrading as soon as possible.

“If you modernize today, you could substantially reduce the number of servers needed to support the same applications at a much lower power consumption rate,” Privett says. He advises sitting down with folks from the OT side of the house to make sure power is available for whatever you want to do. In many areas, power is at a premium.

If your plans include installing GPU environments in your own data center, WWT advises you not to delay. “We’re telling customers, you need to talk with us and get that designed, get that ordered, because it will take quite a bit of time until it actually ships and we’re able to install it,” Anderson says.

Moor Insights’ Kimball agrees. “You have to order these parts today if you want to see them hitting your dock, your warehouse, or your office 12 months from now.”

Dell Technologies is acknowledging that infrastructure and storage supply still can’t keep up with agentic AI’s insatiable appetite for resources.

The company this week reported a “record” AI backlog, with $95 billion in orders waiting to be filled. This dovetails with quarterly earnings reflecting a more than 50% year-over-year increase in AI demand.

On an earnings call, Dell COO Jeff Clarke acknowledged that supply constraints start with servers and storage, and span the stack to “just about every product going through a leading node.”

“We are doing everything we can to get more supply,” he said. “In today’s environment, that’s a very difficult task.”

A glimpse of infrastructure demands ahead

Dell reported that, in its financial quarter ending July 31, its revenue was $47 billion, reflecting 58% year-over-year growth. Moreover, revenue in its Dell Infrastructure Solutions Group (ISG) increased 89% to a record $31.8 billion.

Much of this growth is in servers, notably traditional central processing unit (CPU)-based servers that are increasingly supporting agentic AI workloads. Demand is “exceptionally strong” in this area, with earnings up 122% year-over-year.

Perhaps most tellingly when it comes to the ongoing demand, the company booked nearly $61 billion in AI server orders in the three months ending July 31; all told, over the last 12 months, it has inked more than $130 billion in AI server orders.

Clarke reported that Dell converted $131.7 billion of demand into orders over the last year, and that demand is broadening across enterprise customers, neoclouds, and sovereign cloud providers. To illustrate his point, he noted that the number of customers using Dell AI Factory, the company’s platform built to support AI workflows, has surpassed 6,500, and of those, 3,300 signed on in the last three quarters. Clarke pointed out that, by contrast, it took the company two years to sign on the first 3,200 after debuting Dell AI Factory in May 2024.

“Agentic demand is reshaping the data center,” Clarke said. Inference is “pure demand in our industry.” In fact, Dell anticipates that 3,600 quadrillion tokens will be in use by 2030, representing an 87x increase from today. Further, over that same period, training demand is predicted to grow to 850 zettaflops, a 5x jump.

“Enterprise agentic AI is expected to be the single largest workload by 2028,” Clarke said, and by 2030 will account for 75% of all data center demand.

Enterprises clamor for traditional servers

Dell is seeing a growing trend of customers requiring “meaningful CPU compute capacity” to support AI and agentic workflows. As evidence of this demand, in just its last two financial quarters, it has generated nearly as much revenue from traditional servers and networking as it has in any prior full year in company history.

Most of this growth comes from existing customers accelerating their investments in traditional IT environments to refresh, modernize, and bolster performance, efficiency, and resiliency. Dell anticipates “significant and durable” refreshes ahead, and heightened security and resiliency requirements are also increasing demand.

“AI requires modern, disaggregated architectures that keep data accessible and in motion across compute, storage, and networking,” Clarke noted. It is much more than assembling and delivering components; AI deployments require significant engineering, design, and deployment expertise. Some customer engagements, in fact, require upwards of 50 unique designs as enterprises optimize for workload performance, power, cooling and the data center environment, he claimed.

Enterprises want new servers with more cores, more dynamic random-access memory (DRAM), and more storage. However, the constraints remain the same: “DRAM, DRAM, DRAM, followed by NAND, NAND, NAND [flash memory],” Clarke said. There are “spotty” CPU and disk drive shortages, and constraints all the way down the supply chain, from microcontrollers to drives to transistors.

Large enterprises and multinational corporations across the globe “would prefer to have products now if we had the supply,” he said. “We are supply constrained in the sense of what we can build in any given quarter.”

This has led Dell to plan accordingly and optimize configurations with what “bits and bytes” they do have coming in to maximize outputs, with a focus on “getting it out the door,” Clarke said. There are associated lead times that the company is working through, but they’ve been able to “realize greater shipments.”

“We’ll continue to focus on trying to get more supply, and take the supply we have and optimize the output,” he said.

Reflecting increased need for storage as enterprises prep, manage, and protect huge volumes of data, Dell has also seen strong growth across its PowerFlex, PowerStore, PowerProtect, and PowerVault products.

“Demand remains broad based; enterprises continue to modernize their storage environments as data growth increases the importance of keeping data available and secure,” Clarke said.

How customers respond to shortages

Clarke acknowledged that modernization is driving higher core counts, more DRAM, and more storage. Those configurations “cost more than they did last quarter, and the quarter before, and the quarter before.”

Customers are adjusting to these price increases, he noted, deferring purchases because they are unable to sufficiently flex existing budget dollars. In other cases, enterprises are placing orders further in advance to ensure they have access to constrained supplies. “Large, sophisticated customers are acting, first and foremost,” Clarke said. Some are collaboratively planning with Dell to gain a view of their needs further into the future.

“That is a new phenomenon,” he said. “We are working through this demand environment that’s well ahead of supply, helping customers manage.”

This article originally appeared on Network World.

Every decade or so, a new form of infrastructure becomes the thing that separates economies that compound from economies that stagnate. In the 20th century, it was ports, highways and power grids. Right now, it’s compute. And governments around the world are scrambling to get a piece of it — offering land, tax breaks and power guarantees to a small group of American and Chinese technology companies — without fully understanding what they’re trading away or what they’re actually competing for.

I’ve spent my career designing and building these facilities. Here’s what I see.

What a country is really signing up for

When a government announces it’s attracting a hyperscale data center, the press release usually mentions jobs, digital transformation and becoming a regional tech hub. What it rarely mentions is what the country is giving up and what it will need to sustain the facility for the next 20 years.

A large data center — say, 100 megawatts — needs roughly the same power as a small city. It needs that power reliably, 24 hours a day, with redundancy built in so that a grid fluctuation doesn’t take down critical systems. It needs water, often millions of gallons per month, for cooling. It needs fiber connectivity with multiple diverse routes. It needs a construction workforce that understands raised floor systems, precision cooling, high-voltage electrical distribution and fire suppression. And it needs all of this before a single server is installed.

Most developing countries don’t have this. Not yet. And the gap between “we want a data center” and “we can sustain one” is exactly where deals fall apart, projects stall or facilities get built and then underperform.

The countries pulling away

The United States has roughly 4,000 data center facilities, more than any other country by a wide margin. That number is growing faster than most of the rest of the world combined. The reasons are structural: deregulated power markets in key states, established fiber networks, deep capital markets, a legal system investors trust and decades of operational knowledge in the industry.

China is building at comparable speed but inside a closed system. Its facilities serve Chinese companies under strict data localization rules. For global capital allocators, China is largely a separate game.

The EU is growing but constrained by its own regulations. GDPR and data sovereignty laws mean European data often must stay in Europe, which is creating demand — but also creating friction. Energy costs, permitting timelines and land constraints in Western Europe are pushing investment toward Nordic countries (cheap hydropower, natural cooling) and Central and Eastern Europe (lower costs, EU membership).

Singapore, Australia and Japan are the established APAC anchors. They have the rule of law, the connectivity and the enterprise demand. But Singapore banned new data center construction outright from 2019 to 2022 over resource concerns, and even its 2025 reopening came with strict sustainability quotas that leave hundreds of megawatts of demand unmet. The pressure is redistributing. 

Where developing countries actually stand

India is the clearest breakout story. It has real enterprise demand, a growing hyperscaler presence and government policy actively supporting data center investment — including a 20-year tax holiday for foreign cloud operators announced in the 2026 budget. The challenges are grid reliability and water scarcity in key metro areas — solvable problems, but they require serious infrastructure investment alongside the facilities themselves.

Southeast Asia — Indonesia, Malaysia, Thailand, Vietnam — is attracting genuine capital. Malaysia in particular has moved fast, drawing more than $24 billion in approved data center investment and positioning Johor (just across the border from Singapore) as an overflow market. The risk is that these countries are capturing construction investment and some jobs, but the operational expertise and long-term value is still flowing out.

Sub-Saharan Africa and Latin America are earlier. There is demand — mobile internet penetration is driving real data needs — but the power infrastructure in most markets isn’t ready for hyperscale. What’s viable today is edge computing: smaller, distributed facilities closer to users that don’t require the same power density. This is where early investors are looking.

What developing countries are getting wrong in negotiations

When a government announces it has attracted a hyperscale data center, the story is always the same: jobs, digital transformation, becoming a regional tech hub. What’s missing from that story is the question of who controls what.

A data center is not an economic anchor the way a factory is. A factory transfers skills, builds supplier ecosystems and creates middle-class employment at scale. A data center run by a foreign hyperscaler employs a small local facilities team, sends all operational decision-making offshore and keeps every dollar of the value it generates inside its own balance sheet. The host country gets the electricity bill and the water consumption. The technology company gets the asset.

What countries are actually competing for is not a building. It’s the right to be inside the infrastructure layer that runs the global economy for the next 30 years. That requires a completely different negotiation — one about data rights, local engineering capacity, grid co-investment and long-term operational control. Almost nobody is having that negotiation. They’re haggling over tax rates instead.

The governments that are negotiating well understand this. They’re demanding local data processing requirements, commitments to train and hire local engineers, co-investment in grid upgrades and technology transfer agreements. They’re treating compute infrastructure the way Gulf states treated oil infrastructure in the 1970s — the leverage point is during the negotiation, not after.

The governments that are not doing this will look back in 20 years and realize they subsidized someone else’s infrastructure empire.

What this means if you’re allocating capital

The investment thesis in this space is not “find the next Singapore.” That window has closed. The actual opportunity is in the infrastructure gaps.

Power is the binding constraint everywhere. Companies that can solve reliable, cheap, clean power for data centers — whether through grid modernization, on-site generation or small modular nuclear reactors — are sitting on the scarcest input in the industry. This is where I’d be looking.

Second-tier markets are real. The “big four” US markets — Northern Virginia, Silicon Valley, Dallas, Chicago — are land-constrained, power-constrained and increasingly expensive. Capital is moving to the Midwest, the Southwest and internationally to markets with available power and land. The facilities being built in these markets today are the critical infrastructure of the next decade.

The countries that get the policy right — stable regulation, reliable power, fair contract enforcement — will attract disproportionate capital. The ones that don’t will keep making announcements and watching projects stall.

In the 19th century, the countries that owned the ports controlled trade. In the 20th century, the countries that controlled oil set the terms for industrial growth. Compute is next. The physical layer of AI infrastructure — the land, the power, the cooling, the fiber — is being locked up right now, mostly by a handful of private companies operating across borders with very little accountability to the countries hosting them.

For capital allocators, the opportunity is real and the window is open but not indefinitely. Power solutions, second-tier markets and policy-stable emerging economies are where the uncaptured value sits.

For governments, the window to negotiate from a position of strength is also now — before the facilities are built and the leverage is gone. Once the servers are in the ground, the terms are set.

The countries and investors who understand this in 2025 will look very smart in 2040. The ones who are still thinking about data centers as a real estate play will not.

On top of July’s 30% price hikes across almost all of its product lines, Nvidia is reportedly preparing to raise prices of servers, including those powered by Vera Rubin and Grace Blackwell chips, by 15%, due to skyrocketing memory prices.

That 15% increase for systems being delivered in early 2027, reported by Bloomberg and other business media, is seen as part of a series of expected price hikes throughout AI deployments.

Scott Bickley, advisory fellow at Info-Tech Research Group, said that he sees Nvidia’s move as one that is only passing along its own rapidly increasing costs. But rather than price gouging because of its close-to-monopoly market control, Bickley’s calculations suggest that Nvidia is likely eating some of its costs, and is only passing along a fraction of them to its largest customers.

But not all AI-related costs are increasing; per-token prices appear to be dropping, he said. That gives CIOs a potential strategy to manage costs by pushing approaches that will reduce the reliance on memory.

“The workload cost is going down per token while the underlying hardware and infrastructure costs are going up,” Bickley said. “If you are directly building out your own clusters, this is an automatic uplift to an already egregiously expensive solution. If you are buying your own hardware, you’re going to have to suck it up. You are not going to negotiate your way out of this.”

But, he added, CIOs should also be able to get more mileage out of the clusters they are currently running, via techniques such as model routing, compression, and batch processing.  

Gaurav Gupta, VP analyst at Gartner, noted that the Nvidia price hikes are reflective of the many pricing increases throughout the AI environment. 

“Memory prices are going up, especially HBM and LPDDR5, but there are other aspects, like leading-edge foundry wafers, advanced packaging, and other component shortages,” Gupta said, adding that those issues generate “longer lead-times, which typically translates to higher prices.” And he does not expect the situation to improve any time soon. 

“In the current environment of strong demand and limited supply, we expect this situation to continue in the near to mid-term,” he said. “This means higher costs for those deploying these servers/systems and for those renting compute in the cloud, including software vendors/model builders, and others.”

Flavio Villanustre, CISO for the LexisNexis Risk Solutions Group, agreed.

“This is a supply and demand problem, and it’s likely to reach a plateau and eventually improve once memory production ramps up to meet the current demand due to AI, but I don’t think this will happen in the next few months,” Villanustre said. “For now, CIOs will need to contend with the current market conditions.”

But he also agreed with Bickley’s suggestion that CIOs try to squeeze more value from the RAM they already have.

“Some AI model vendors are adapting their models to run better in memory constrained environments,” Villanustre noted. “For example, Gemma E4B and similar models by Google now use a hierarchical tiered model that allows them to pull only the necessary parts of the model into memory instead of holding the entire model in RAM. These models have a much larger effective number of parameters than the memory that they require.”

And, added Mike Wilkes, enterprise CISO at Aikido Security, this means that the Nvidia price hikes may do some good if they convince enterprises to adopt a more thoughtful approach to AI deployments. 

“Enterprises have spent the last few years treating frontier-model tokens almost as an infinitely elastic utility, sending workloads to the biggest model whether or not the task required frontier-level reasoning,” he said. “Higher infrastructure and token costs should force much better workload discrimination.”

He observed that the right enterprise AI architecture is increasingly hybrid: reserve 10% for frontier model consumption for problems that genuinely require it, while pushing classification, extraction, summarization, routine agent actions, and other bounded workloads toward small language models (SLMs) and open-weight models running on infrastructure that the enterprise controls.

“That gives CIOs leverage against price gouging or unilateral price setting,” he pointed out.

This article originally appeared on NetworkWorld.

A single camera, sensor, or access control device is unlikely to attract much attention in an enterprise sustainability strategy. It consumes relatively little energy. It operates quietly in the background. It represents only a fraction of an organization’s overall technological footprint.

But when you multiply that device by thousands—or even the tens of thousands—then the calculation starts to look very different.

That’s the reality of today’s enterprise edge. Connected devices are embedded throughout buildings, campuses, stores, factories, and other environments. The purpose is admirable; they’re intended to support security, safety, and day-to-day operations. And since many devices operate around the clock and remain in service for years upon years, they create a cumulative energy footprint that can be easy to overlook.

For CIOs motivated to reduce emissions and operational costs, that makes the edge an important part of the sustainability equation. Cloud efficiency and data center infrastructure are already important parts of that conversation, but they don’t provide the full picture. Organizations also need visibility into the devices operating at the edge.

Axis Communications analyses show that between 60% and 80% of a network camera’s total environmental impact comes from the energy consumed during the use phase of the device. This is primarily because these devices are designed to operate consistently over several years, leading to energy consumption that outweighs the impact of material extraction, manufacturing, and logistics. 

For CIOs facing increased accountability around  Scope 2 and Scope 3 emissions, this fact presents both a challenge and an opportunity. While many decision-makers lack a clear understanding of how much energy these systems actually consume, that uncertainty is compounded by the thousands of connected edge devices organizations manage. Ultimately, you can’t optimize what you don’t measure, and without the clarity, sustainability reporting and strategies remain incomplete. 

There is good news via a modern solution. Smarter edge technology can significantly reduce environmental impact while also improving a business’s operational efficiency. 

Modern edge devices are designed to do more with less. Developments in intelligent software, hardware, and advanced compression reduce the amount of data that needs to travel across networks or be stored in power-hungry devices. Additional features like high-efficiency SoCs (System on a Chip) and advanced compression can further reduce downstream energy demands tied to bandwidth and storage. 

The result? Measurable sustainability gains that don’t sacrifice performance or security. 

Efficiency matters at scale 

“If you take one of our P32 cameras, the energy consumption during normal use is less than a regular LED light bulb,” says Ulrika Renmark, Sustainability Sales Engagement Director at Axis Communications. “The energy consumption of servers, on the other hand, is measured in hundreds of watts. And if you include the air-conditioning of the server room, then we are talking thousands of watts.” 

Renmark points to Zipstream, an Axis-developed optimization technology, which reduces bandwidth and storage needs without sacrificing image quality. Similar innovations can significantly reduce the energy footprint associated with supporting infrastructure, and they represent what responsible edge innovation should look like in today’s high-tech world. 

This shift reflects a broader opportunity in enterprise IT. Sustainability gains increasingly come from intelligent architecture decisions rather than standalone sustainability initiatives. Processing data at the edge, where appropriate, can reduce unnecessary cloud traffic, lower latency, and improve resilience, all while minimizing energy consumption. 

Hybrid architecture is especially well-positioned to support these goals. By balancing processing between edge devices and the cloud, organizations can optimize performance, security, and sustainability. This approach contributes to improved device lifespan, improved security posture, and reduced operational waste. 

Cloud-enabled lifecycle management can further amplify these opportunities. Remote diagnostics, firmware updates, and centralized device management reduce the need for on-site maintenance visits and manual interventions. That not only cuts operational costs but also lowers emissions associated with travel and hardware replacement. 

The path forward starts with visibility 

In their quest for sustainability, organizations should begin by auditing edge infrastructure to identify where energy consumption is concentrated across devices. Procurement teams should prioritize vendors that provide transparent energy data and demonstrate continuous improvements in efficiency. Sustainability considerations should also extend beyond energy use to include repairability, lifecycle management, and total cost of ownership. 

The ultimate goal isn’t to necessarily make edge infrastructure the centerpiece of an organization’s sustainability strategy. It’s to make sure edge infrastructure isn’t missing from the conversation.

As mentioned, a single device may represent only a small opportunity for improvement. Across thousands of cameras, sensors, access control systems, and other connected technologies, however, the incremental sum in energy consumption can add up.

As enterprise technology environments continue to expand, sustainability will increasingly depend on decisions made throughout the entire infrastructure. Understanding the impact of those decisions will help organizations find efficiencies that might otherwise remain hidden.

To learn more, visit us here.

Oracle said it will deploy Quantinuum’s Helios quantum computer inside its cloud infrastructure and provide enterprise customers access through a planned Oracle Cloud Infrastructure (OCI) quantum service for hybrid quantum-AI workloads.

The companies have forged a multi-year partnership in this regard.

The partnership will see Helios installed in a US-based OCI AI data center, where it will support hybrid quantum-AI workloads delivered as a cloud service. Oracle said it plans to preview the OCI quantum service in the coming months.

“With Quantinuum’s Helios on OCI, customers can expect to gain managed, secure access to cloud-hosted quantum computing without having to procure, install, or operate dedicated hardware or specialized facilities,” the companies said in a joint statement.

OCI to deliver hybrid quantum-AI workloads

The partnership is focused on enabling hybrid workloads that combine quantum computing with classical HPC and AI systems, according to the statement.

Under the arrangement, OCI customers will be able to access the Helios system alongside existing GPU and HPC resources.

“By operating on-premises within OCI’s infrastructure, Helios is anticipated to be able to integrate seamlessly with existing OCI compute, networking, storage, identity, and data services under the same governance and access controls customers already use,” the statement added.

By hosting Helios within OCI infrastructure, customers will be able to access quantum resources without procuring, installing, or operating dedicated hardware or specialized facilities, the statement said.

Oracle said the planned OCI quantum service will allow developers to move from simulation to execution on quantum hardware. The service is expected to combine Quantinuum’s development stack with support for open-source hybrid programming frameworks, enabling developers to build and test quantum-classical applications.

Charlie Dai, vice president and principal analyst at Forrester, said the integration lowers access barriers by embedding quantum computing into existing enterprise environments. “Integrating Helios into OCI lowers access barriers by embedding quantum into existing cloud governance, security, and AI/HPC workflows,” he said.

However, Dai added that this shift does not materially change adoption timelines. “Most organizations remain in the PoC phase, with production quantum advantage still limited to a narrow set of problems,” he said.

Use cases center on research, early enterprise exploration

The companies said the platform could support applications across drug discovery, materials science, financial modelling, and large-scale optimization, including AI workloads.

“Deploying Helios inside OCI gives Quantinuum and Oracle an opportunity to create a unique and deeply integrated environment for hybrid workloads, explore enterprise use cases with customers, and accelerate commercial adoption,” Quantinuum President and CEO Dr. Rajeeb Hazra said in the statement.

The statement added the partnership will explore enterprise, AI lab, academic, and research applications, and reflects a shared view that combining quantum computing with AI and classical systems could help address computational challenges that are difficult to solve using conventional approaches alone.

Dai said CIOs should approach hybrid quantum-classical workloads as a long-term capability-building exercise rather than a near-term ROI driver. “For most enterprises, the business case today is capability building rather than measurable operational ROI,” he said.

He added that near-term value is likely to come from areas such as algorithm development, workforce readiness, and targeted research in optimization, materials science, and drug discovery.

Enterprise adoption remains focused on experimentation

Helios, Quantinuum’s third-generation quantum computer, is based on a 98-physical-qubit trapped-ion architecture and has been used in demonstrations involving 48 logical qubits, the statement said.

Quantinuum said the Helios system is designed for hybrid integration with classical HPC and AI environments.

The statement said a single Helios system has an estimated power draw of less than one percent of the draw reported for leading supercomputers, positioning it as a lower-power complementary resource for suitable hybrid workloads.

Dai said Oracle’s move aligns with broader quantum-as-a-service strategies from hyperscale cloud providers.

“Oracle’s move aligns with broader quantum-as-a-service strategies from other hyperscalers,” he said, adding that while the integration strengthens the developer environment, enterprise adoption remains focused on experimentation rather than broad production deployment.

The article originally appeared on NetworkWorld.

OVH is increasing the prices of its servers, some by as much as 87%, for both new and existing customers, blaming AI’s insatiable demand driving the rising cost of the RAM and storage it uses in its data centers.

The European cloud operator specializes in low-cost bare metal and public cloud offerings.

CIOs will be familiar with the balancing act OVH has had to perform over the last year. In a Monday post explaining the upcoming increases, OVH chairman Octave Klaba wrote on X,  “We have to place the right volume of orders, month by month, over 12 months, with no guarantee of the purchase price and without knowing what will be the real demand from our customers.”

Still, he added, “even though our prices are increasing, we remain the cheapest on the market for bare metal and public cloud; where before we could be 3x cheaper, we will be 2x cheaper (if our competitors don’t increase their prices).”

The increases will hurt hard-core gamers hardest, with the cost of the company’s most recent gaming servers rising 87%. (Older gaming instances are unaffected.)

High Grade, high price

But enterprises will also feel the pain from climbing component costs: OVH’s latest High Grade bare metal servers, with up to 2 x 96 cores of AMD Epyc 9005 series processors, 36 hard disks per server, and high-density cooling systems, will go up in price by 59%; older models built to the 2024 spec will go up 26%.

Lower-performance servers will also see increases of 40%-49% for the most recent models, and 26%-37% for older models.

The new prices take effect from Sept. 1 for new orders, and from Oct. 1 for renewals.

It’s not just baseline server prices that are increasing; optional additional memory and storage are going up in price too. OVH already increased the cost of these extras for new server orders as of July 1, with RAM prices rising 127% and disks 89%. From Oct. 1, renewals will be affected too, with the price of additional RAM in the latest servers rising by 40%, and that of larger disks by 15%. For servers built to 2024 specs, the increases will be 20% and 10% respectively.

Existing customers can lock in current prices for servers already in production for up to four years if they pay in advance by Oct. 1, Klaba wrote. Existing commitments will not be affected by the increases until they are due for renewal.

Small instances, big increases

The price rises are more nuanced when it comes to public cloud systems. In future, OVH will break out storage and IP address rental costs separately, and will allow customers to mix and match storage capacity and compute.

“In appearance, hourly compute cost won’t change,” Klaba wrote. “On the other hand, low-latency Block Storage and IPv4 addresses, previously included in our Gen3 instances (B3, C3, R3) will appear as two separately billed line items on Oct. 1.”

The result is price increases of as little as 1.4% for the most powerful instances, or as much as 21.9% for smaller instances, he said.

OVH will continue to offer a 15% discount for a commitment of one year, or 30% for three years, he said, but will no longer offer discounts for shorter terms.

This article originally appeared on NetworkWorld.

As enterprises look for ways to cut the cost of running AI models in production, AMD is betting that not every AI workload will be best served by a power-hungry general-purpose GPU.

AMD has agreed to buy Taalas, the Canadian designer of chips that permanently embed a trained AI model’s weights into custom silicon, instead of repeatedly loading them from memory during inference as conventional GPUs do.

Taalas says its approach reduces the time and power required to move model weights between memory and compute units, making things run faster and cheaper.

The result is a highly specialized inference processor optimized for one model, trading the flexibility of programmable hardware for substantially higher throughput and energy efficiency.

Operational tradeoffs

While AMD is planning to integrate the chips into its Instinct GPU roadmap, targeting system-level AI inference solutions in data centers, analysts remain skeptical that enterprises will readily embrace hardware tied to a specific AI model.

Enterprises would, effectively, be buying a chip and a model together because unlike GPUs, which can be repurposed to run different AI models through software updates, Taalas’ chips are tied to a specific trained model, meaning they would need different hardware to support different inference tasks, said Amit Kumar Jena, AI development manager at IT Consulting firm Kanerika.

Or as Forrester Principal Analyst Charlie Dai put it, “The biggest risk is inflexibility.”

The requirement to swap hardware in order to swap tasks would, Dai said, introduce new challenges with costs, governance, capacity planning, lifecycle management, and supplier dependency, especially for enterprises managing multiple AI workloads.

Manoj Chandra Jha, principal analyst at Nord-IQ Research, said the risk of fusing chip and model into one component is larger than one might think, as “early model obsolescence strands both together, so this should be modeled as one shorter-lived asset rather than two independently amortized ones.”

Taalas says it can update a model by modifying only two metal layers of the chip rather than redesigning it from scratch, but that will only apply to chips that haven’t yet left its factory, not those already in use.

That means enterprises will still need to plan for hardware refresh cycles measured in weeks or months and retain programmable GPUs for workloads that evolve frequently, said Pareekh Jain, principal analyst at Pareekh Consulting.

It also means, said Jha, that what is typically a software decision becomes one about capital expenditure for Taalas customers, as replacing or switching workloads or models could require investing in new hardware rather than simply updating software.

Where model-specific silicon fits

Those tradeoffs significantly narrow the range of enterprise workloads where model-specific silicon is likely to make economic sense.

Dai sees the technology as best suited for mature, predictable inference workloads that run at massive scale and rely on relatively stable AI models, such as customer service automation, fraud detection, industrial computer vision, network operations, edge AI, and embedded copilots.

For CIOs, that effectively limits model-specific silicon to a small subset of enterprise AI deployments, rather than a wholesale replacement for GPU infrastructure, he said. “GPUs will remain the preferred enterprise platform because most enterprises value flexibility, multi-tenancy, and rapid model evolution over maximum efficiency.”

This article first appeared on Network World.

A likely US administration ban on Chinese optical transceivers for AI data centers may have an unintended consequence: IT will rush to buy as many of the components as possible before restrictions kick in.

The US Federal Communications Commission (FCC) “is working on the measure to bar imports of new Chinese optical transceivers” and officials hope to publish and implement it this year, Reuters reported on Tuesday.  

The report, citing four sources familiar with the matter, said that the official rationale is “to prevent Chinese firms from stealing data, installing malware or disrupting service at US data centers.” The sources did, however, stress that such a ban could still be modified or shelved.

A valid concern

Analysts and consultants agree that the concern, albeit hypothetical at the moment, is valid. 

If implemented, such a ban would have a severe impact on data center (DC) strategies for both enterprises and hyperscalers. Although higher costs for replacement products would be all but certain, the greater concern is the lack of availability of non-Chinese transceivers and other components, regardless of price. 

A potentially even more worrying element of a ban is the need for far more sophisticated supply chain visibility. That is because many of those non-Chinese component suppliers actually use some Chinese components in their products, which means that the exact wording of any potential FCC restrictions will be critical. It will define how closely enterprises will need to examine their suppliers’ supply chains.

Aman Mahapatra, chief strategy officer for Tribeca Softtech, a New York City-based technology consulting firm, said that he thinks that an FCC ban is quite likely, because it “has run this exact playbook four times in eighteen months, against drones, routers, robots, and the July 28 inverter and robotics restrictions. The mechanism is tested, the machinery is warm.”

If the ban is enacted, said geopolitical analyst Irina Tsukerman, “CIOs will need to reassess vendor diversification, and other factors such as replacement compatibility, lifecycle planning and inventory management, given that many organizations have historically treated optical components as interchangeable commodities.”

“Enterprises will also need much greater visibility into firmware development, manufacturing origin, as well as subcontractors, and software update processes, because future procurement decisions are increasingly likely to examine the entire supply chain rather than simply the company selling the finished product,” she added. This will make future procurement more complex.

IT pain will vary

Tsukerman said that, although prices would certainly spike, the pain felt will vary based on the nature and size of each affected business She noted that while hyperscale operators can negotiate directly with manufacturers, secure long-term supply, and qualify multiple vendors for critical components due to their purchasing power, enterprises, regional data center operators, and colocation providers generally lack that leverage. Rather, they often depend on distributors supplying lower-cost Chinese products, making them considerably more vulnerable to price increases and delivery delays.

Mahapatra added that the preliminary indications suggest any such ban would have a “new models only” framing that would protect the installed base while restricting the next generation of products, which, he said, would be a compromise “generous enough to mute the hyperscaler objection.”

But, he said, “the enterprise CIO running a colocation expansion or private AI cluster is about to discover they are competing with Microsoft and Meta for the same constrained supply and losing.”

He recommended that enterprises lock down forward optics supply for anything they plan to build through 2028 before the restriction publishes, because, he pointed out, “announced-but-not-effective bans consume non-Chinese capacity through panic buying, and buyers who move after publication pay in schedule rather than dollars.”

However, such a move depends on how serious IT considers the cybersecurity risks from the Chinese components. Tsukerman argued that data leakage and malware fears need to be taken seriously, because modern optical transceivers often contain firmware, onboard memory, and management interfaces, and may also offer capabilities that can influence how traffic is monitored and managed throughout the data center.

In that case, she noted, “the risk would extend beyond espionage to include compromised firmware updates, manipulation of diagnostic information, disruption of maintenance support, delayed replacement shipment, or in the worst case scenario, interference with critical infrastructure during periods of heightened political tension.”

However, Mahapatra sees the risk quite differently.

“A transceiver is a comparatively dumb device converting electrical signals to optical and back, not a router running a network operating system with deep packet visibility,” he said. “The near-term espionage risk from currently shipping Chinese optics is thin, and CISOs who reallocate budget toward this threat over their software supply chain and identity attack surfaces are responding to headlines rather than risk.”

The suppliers involved

Consultant Brian Levine, executive director of FormerGov, labeled the potential US administration move as “one of the more consequential supply-chain moves the FCC has contemplated, because optical transceivers are the workhorse components that move data across fiber inside every AI data center, and Chinese vendors dominate that market.”

He noted that Chinese vendors Innolight and Eoptolink alone reportedly account for the majority of the 800-gig modules going into Nvidia’s AI clusters, so a ban “wouldn’t be a minor substitution,” and non-Chinese alternatives such as Coherent and Lumentum in the US don’t yet have sufficient capacity to fill the gap.

Nader Henein, a Gartner VP analyst, agreed, adding that since the nature of the AI data center supply chain is both complex and fragile, a small change can deliver a disproportionate industry impact.

“If you remove one provider from the equation, it’s not as if the others have capacity to fulfil the increase in demand, so it’s not simply a question of added cost, it’s a question of placing a ceiling on capacity and growth,” he said.

Tsukerman said that her list of the companies most likely to benefit from such an FCC ban would include Coherent, Lumentum, Applied Optoelectronics and Cisco’s Acacia business, while Broadcom and Marvell, as well as  Japanese and Taiwanese manufacturers, also provide important optical and connectivity technologies that support advanced networking infrastructure.

Other components in the crosshairs

She pointed out that there is also a strong probability that a transceiver ban would quickly be followed by attacks on other components. 

Networking switches, SmartNICs, data processing units, baseboard management controllers, storage controllers, intelligent power distribution units, cooling management controllers, optical transport systems, and embedded management processors “all perform functions that could influence the operation of an entire facility if compromised,” she said. “None of these products simply passes data or delivers electricity. They manage, monitor, or control critical infrastructure, making them increasingly attractive targets for supply-chain attacks.”

Her list of likely future US targets for restrictions also includes top-of-rack switches, spine switches, and rack management systems,.

Flavio Villanustre, CISO for the LexisNexis Risk Solutions Group, echoed Henein’s fears about industry impact.

“I think that the appropriate response to these types of risks needs to be more nuanced than just a blanket ban,” he said. “Since 15%-20% of all world’s semiconductors are manufactured in China, and that number rises to 80% or 85% if you include Taiwan, blocking Chinese imports for these components could hamper the entire datacenter industry.”

Although there have been rumors of insecure or trojanized hardware components sourced from China in the past, given that many large American and multinational technology vendors manufacture their parts there, ”a sledgehammer approach could spike prices for these types of systems, jeopardizing development of new technologies,” he noted. “A far more reasonable approach would be to require appropriate testing and quality controls to ensure that those risks are appropriately mitigated.”

Would likely harm the US

In addition, independent technology analyst Carmi Levy said that he is skeptical about whether an FCC ban would ultimately be a good move for the US.

“It’s fair to ask whether this will truly make American technology infrastructure more secure, or whether it’s little more than a performative stunt designed to score geopolitical points,” Levy said, pointing out that Canada didn’t end up any safer because of the Huawei and ZTE ban, and “no one should fool themselves into believing a Chinese data center ban in the US would be any different. It would only add further constraints to a supply chain that’s already close to collapsing under its own weight [and it] will likely harm American interests more than anyone else’s.”

But he also concluded that such a move would likely fail, given the current global state of data center technologies. 

“Chinese suppliers and components have been so ingrained in the global technology supply chain for so long that no ban of any form could hope to have any tangible impact on so-called national security,” Levy said. “To claim otherwise exposes the true motivations of this misdirected policy strategy.”

This article originally appeared on NetworkWorld.

A growing backlash against building new data centers in the US may have huge cost implications for CIOs planning to expand their organizations’ AI initiatives.

Protests against building new data centers were organized in 42 states in mid-July, with participants concerned about new facilities driving up electricity and water costs and using large swaths of land.

As of mid-July, 10 states, including Florida, Georgia, and Virginia, had active data center construction moratoriums in place, and eight other states had pending legislation, according to datacenterbans.com.

In addition, as of May, 23 states had approved large-load tariffs that require data centers to pay the full infrastructure cost for their facilities, says Arif Gasilov, a partner in the natural resources and built environment division of sustainability advisory firm Gasilov Group.

IT leaders need to calculate the backlash into their planning for the compute and other IT infrastructure needs that new data centers would meet, he says.

“What this means for CIOs is that power cost assumptions built in 2023 are wrong in close to half the country,” Gasilov says. “A CIO planning an AI deployment that depends on colocation or cloud capacity in any of these states should be asking their provider what the rate structure looks like under the new tariffs and recalculating economics.”

In some cases, it may be possible to go smaller to avoid the moratoriums or tariffs on large data centers, but some state regulations target facilities close to each other as opposed to individual data centers, he notes.

Deployment challenges

If the backlash continues, IT leaders may need to rethink the way they deploy AI, says Chuck Girt, CTO at fiber-optic network provider FiberLight.

With fewer options for AI compute power, organizations would have less flexibility in where they deploy AI workloads, he suggests.

“I don’t think the rate of data center construction changes the direction AI is headed, but it could influence how organizations deploy and access AI at scale,” he says. “Most enterprises aren’t going to build this infrastructure themselves; they’re going to rely on cloud and data center environments to provide the compute AI requires.”

A lack of data center options could put many organizations in a bind, says Kevin Surace, CEO of biometric security vendor TokenCore.

“Compute capacity is becoming as strategically important as electricity, semiconductors, and network connectivity,” he says. “Fewer data centers mean less available capacity, reduced geographic redundancy, longer provisioning times, and greater dependence on a small number of cloud providers and locations.”

Organizations that have not secured capacity could find that their AI strategy is technically sound but physically impossible to execute on schedule, he suggests.

Surace, also an AI and green energy expert, is concerned that generalized fear about older data center designs is turning into blanket opposition to new construction. Modern facilities have cut down on the massive water use of older data centers, he notes, and some are using renewable energy generation. Nuclear power will become an electricity option soon, he adds.

Cost pressures rising

In the meantime, IT leaders should expect higher costs for compute and other IT infrastructure provided through data centers, Surace says.

“Demand for AI compute is accelerating, so constraining the supply of facilities, electricity and high-density capacity will place upward pressure on cloud pricing, colocation, accelerator access, and long-term capacity contracts,” he adds.

Organizations that have the capacity will should be able to protect themselves through multiyear agreements and dedicated infrastructure, he suggests. Smaller organizations, startups, and universities could face the greatest percentage increases and may simply be priced out of leading-edge AI capabilities, he adds.

Therefore, Surace advises CIOs to treat compute and energy as strategic supply-chain risks. Organizations should secure capacity as soon as they can, avoid dependence on one cloud or one geographic region, and use smaller and more efficient AI models where appropriate, he recommends.

He also suggests that CIOs ask data center providers several hard questions:

• Where does the water come from?
• Is the cooling loop closed?
• Who pays for new grid infrastructure?
• What percentage of power is generated onsite?
• What environmental monitoring is publicly reported?

Data centers can mitigate some of the community concerns, he says. “Transparency and early community engagement are far less expensive than lawsuits, project cancellations, and moratoriums,” he adds.

Backlash against inefficiency

While protests are likely to continue, some don’t see the concerns about data centers as a condemnation of AI. Instead, the problem is with inefficient AI deployments, says Anurag Gurtu, cofounder and CEO of agentic AI platform provider Airrived.

“Enterprises don’t actually want more data centers; they want more intelligence per watt, per GPU, and per dollar,” he says. “The winners won’t be those with the biggest infrastructure footprint, but those extracting the most value from every unit of compute.”

Limitations on data centers will impact companies only if their AI strategies depend on nearly unlimited infrastructure, he adds.

“The next generation of AI will be constrained by compute, power, and economics,” Gurtu says. “Organizations that optimize models, deploy domain-specific AI, and leverage hybrid architectures will continue to innovate, while those relying solely on scaling hardware will face diminishing returns.”

While limited compute options could lead to higher prices, the solution is to focus on efficiency, he adds.

“Rising infrastructure costs also accelerate innovation in model optimization, inference efficiency, and intelligent orchestration,” Gurtu says. “History shows constraints often become the catalyst for the next wave of breakthroughs.”

Oracle on Tuesday described new EBCDIC character set compatibility features in Oracle AI Database for customers transitioning from legacy databases on IBM mainframes.

In its post, Oracle noted that EBCDIC compatibility has historically been one of the top technical challenges for enterprises re-platforming to use newer, ASCII-based databases while continuing to use proven legacy applications.  

“Achieving this goal requires more than simply moving data. It requires preserving the EBCDIC compatibility on which existing applications depend,” wrote Michael Yau, VP for Oracle Database Globalization Engineering. The feature rollout “addresses two fundamental challenges of preserving EBCDIC compatibility: accurate character encoding conversion and preservation of EBCDIC binary ordering.”

He added: “These client character sets implement IBM Character Data Representation Architecture (CDRA) code page definitions, providing source-to-target character mappings that are compatible with IBM’s published standards. This enables accurate and predictable character encoding conversion during data migration and subsequent database client/server communication.”

Yao observed that this is important because these mainframe migrations can be very complex.

“Many legacy EBCDIC applications, such as those written in COBOL, implicitly rely on the EBCDIC binary ordering defined by IBM EBCDIC code pages. SQL predicates that compare character values, perform range searches, or sort query results often assume this ordering,” he wrote. “After migration to an ASCII-based Oracle AI Database character set, these same SQL statements can produce different results, not because the data changed, but because the database’s default binary ordering follows that of the ASCII-based database character set rather than the source EBCDIC code page.” 

Compatibility repair, not modernization

While consultants generally applauded the new features, some questioned whether this will simply shift enterprise dependency on IBM to dependency on Oracle. 

Sanchit Vir Gogia, chief analyst at Greyhound Research, is one of the fans.

“Oracle has repaired one of the oldest silent faults in mainframe migration: EBCDIC ordering, the muscle memory of the legacy estate. Preserve the data and lose the ordering, and a query returns the wrong record while every dashboard stays green,” he said. “The application runs and the query completes. The answer is simply wrong.”

Gogia noted that the new feature is “not a modernization suite. It is a compatibility repair, narrow and genuinely useful, which CIOs who have bled on past migrations will read with equal parts relief and suspicion.”

AJ Thompson, CCO at UK IT consulting firm Northdoor, agreed that the Oracle announcement addresses a genuine technical barrier rather than just being a marketing gimmick, so it is worth taking seriously as a re-platforming enabler. “The two problems it solves, EBCDIC to ASCII character conversion and preserving EBCDIC binary sort ordering, have historically been real blockers for allowing COBOL to move away,” he said.

But, he cautioned, CIOs must also take resiliency challenges seriously. “Mainframes are not chosen primarily for character encoding, they are chosen for benefits like decades of proven uptime, IBM Z’s redundancy architecture, and workload isolation,” he pointed out.

“Oracle’s announcement solves a data compatibility problem, not a resilience or availability one. A client with genuinely mission-critical, zero-downtime workloads will still need convincing on the availability and disaster recovery side before moving [to the cloud], and Oracle’s own resilience claims would need scrutiny on their own merits, quite separate from this EBCDIC work.”

Leverages IT desperation

Mike Wilkes, enterprise CISO at Aikido Security, added that Oracle is leveraging IT desperation to squeeze long-term value from legacy systems. 

“I have always believed that Oracle will own the very last white-knuckle-grip workloads that migrate from on-premises data centers into the cloud,” he said. “This announcement certainly demonstrates that they understand their position in the world of cloud service providers. They are not the biggest, they are not the oldest, and they are not the most technically advanced. But they do own the market share for the trailing edge of cloud adoption.”

He pointed out that the greatest barrier to cloud migration is not containerizing modern applications, it is the decades of business logic buried inside COBOL applications and EBCDIC-encoded data.

“Oracle’s EBCDIC compatibility features acknowledge a practical reality: organizations are not rewriting these systems from scratch,” he said. “If Oracle can reduce the cost, risk, and operational disruption associated with moving those workloads, the announcement represents meaningful value for IT teams that have been delaying modernization because the migration path was simply too complex or too expensive.”

Could cause vendor lock-in

Then again, Wilkes noted, there is the potential for increasing vendor lock-in.

“Compatibility layers almost always increase long-term dependence on the platform providing them,” he said. “Rather than eliminating legacy technology, they abstract it behind Oracle’s database and cloud ecosystem, making future migrations potentially more difficult. Enterprises should view these tools as transition accelerators rather than permanent architecture.”

Thus, he said, if they use the opportunity to gradually modernize applications and data models, there is substantial value, but if they simply relocate technical debt into Oracle Cloud, they may find that they have exchanged one form of legacy lock-in for another.

Easier, but not easy

Matt Kimball, VP and principal analyst with Moor Insights & Strategy, also saw the Oracle move as a good one, but stressed that while it should make things easier for organizations, re-platforming still won’t be easy because EBCDIC migration isn’t just a character-conversion exercise. However, “Oracle’s built-in character-set support and EBCDIC collations move part of that compatibility burden into the database, making it easier to preserve existing application behavior,” he said.

Ishraq Khan, CEO of coding productivity tool vendor Kodezi, agreed. 

“One of the biggest obstacles to leaving mainframes is decades of applications built around EBCDIC encoding and legacy data formats. If these compatibility features reduce the amount of code that needs to be rewritten, they can lower migration risk, cost, and implementation time,” Khan said. But, he added, organizations should also consider whether these features simply make migration easier, or make future moves more difficult.