Many companies that set out to build on-premises AI infrastructure are now hitting an unexpected wall. Storage budgets drawn up just months ago no longer add up once procurement actually begins—prices have surged that dramatically. According to an Omdia survey commissioned by Verge.io (400 North American IT professionals in storage and infrastructure roles, conducted May 6–26, 2026), DRAM prices nearly doubled in Q1 2026, while SSD prices rose more than 50%. 98% of respondents said they were either considering or already implementing storage tier consolidation, and among companies that hadn't yet adopted it, the technology cited most often as a top investment candidate was software-defined storage (SDS). Now that hardware pricing has spiraled out of control, companies are rethinking the very design philosophy underlying their storage foundations.
DRAM Doubles, SSDs Up 50%: Omdia Survey Reveals North American IT's Distress
According to the survey commentary published by Verge.io, Omdia's characterization of "nearly doubled" and "up more than 50%" is based on TrendForce data showing that DRAM contract prices rose 90–95% quarter-over-quarter in Q1 2026, while enterprise SSD contract prices rose 53–58% over the same period. The overall NAND flash contract price outlook was also revised sharply upward, from a projected 33–38% increase as of January 2026 to 85–90% by March—multiple data sets pointing in the same direction of a sharp price spike.
In response to these price swings, 98% of survey respondents said they were either considering or already implementing storage tier consolidation. When companies that had not yet adopted SDS were asked which of eight technologies they were considering investing in, SDS was chosen most often. Simon Robinson, Chief Analyst at Omdia, commented, "We were surprised by how many organizations are looking at software-defined storage as a potential solution." The fact that SDS came out on top reflects not only a direct response to soaring prices, but also the perception that it offers a practical option for extending the life of existing storage investments without a major overhaul.
Among the same 400 respondents, roughly three-quarters said rising storage costs were causing them to slow down on-premises AI deployment. Moving to the cloud is an option, but for companies in regulated industries or those managing large datasets, retreating from on-premises infrastructure isn't a decision that can be easily reversed. This is the underlying structure at play: storage budget reviews are directly delaying AI deployment plans themselves.
Why Do SSD Prices Keep Climbing? A Production Shift Among NAND Makers
At the root of the price surge is a shared "profit-first shift" among memory manufacturers. Samsung and SK hynix, which produce both DRAM and NAND, have prioritized their limited manufacturing capacity for HBM (High Bandwidth Memory—a component made by stacking multiple DRAM chips for higher speed) production in response to surging demand from AI accelerators, relatively curtailing investment in general-purpose DRAM and NAND. According to semiconductor industry reports, both companies plan to reduce their 2026 NAND wafer output compared to 2025 (Samsung from 4.9 million wafers per month to 4.68 million; SK hynix from 1.9 million to 1.7 million). The decision to cut production in a rising market has pushed supply and demand even tighter.
Kioxia and SanDisk, which only manufacture NAND and cannot produce HBM, have found themselves filling the NAND supply gap left by Samsung and SK hynix. At the same time, NAND manufacturers—including these two companies—are increasingly prioritizing production of enterprise SSDs for data centers, which carry higher margins than general-purpose products. According to semiconductor industry reports, the five companies Samsung, SK hynix, Kioxia, SanDisk, and Micron have nearly sold out their 2026 NAND production capacity. While some new facilities, such as Kioxia and SanDisk's Kitakami plant, began operations in the latter half of 2025, it typically takes several quarters of phased ramp-up before new investment translates into meaningful production volume reflected in supply. There is no prospect of supply rapidly expanding any time soon.
The business logic—that redirecting production capacity toward high-value-added AI products yields higher margins than selling low-margin general-purpose products in bulk—is rewriting the industry's entire supply structure. As a result, the supply of general-purpose enterprise SSDs that companies have relied on for on-premises storage procurement has grown relatively scarce, and prices have risen ahead of everything else.
The results show up clearly in the financial figures. According to data published by TrendForce on June 11, 2026, the enterprise SSD industry's Q1 2026 revenue reached $18.5 billion (approximately ¥2.9415 trillion at ¥159 to the dollar), up 86.1% quarter-over-quarter. The breakdown was Samsung at $7.05 billion (up 92.8% QoQ), the SK hynix group at $4.64 billion, Micron at $3.09 billion, Kioxia at $2.22 billion, and SanDisk at $1.47 billion. The 53–58% rise in contract prices mentioned earlier translated directly into revenue growth for each manufacturer. What's a tailwind for sellers is a headwind for buyers.
What Is Software-Defined Storage (SDS)? Explaining How It Works
SDS refers to a system in which storage control functions—such as data placement, replication, and tiering—are separated from dedicated hardware and run at the software layer. Traditional storage typically came as vendor-specific appliances that bundled hardware and control software together. SDS instead runs these control functions as software on general-purpose servers, allowing disks from different generations and different vendors to be treated as a single pool.
This approach works because of a simple fact: data naturally splits into "frequently used" and "rarely used" categories. Major SDS products such as VMware vSAN and Weka include features that automatically tier data based on access frequency, allowing companies to place only truly needed hot data on fast but expensive new SSDs while moving infrequently accessed cold data to existing, cheaper HDDs. Rather than buying more SSDs at current inflated prices every time capacity runs low, this creates room to reuse older-generation hardware already on hand—which is exactly why it's drawing attention amid the price surge.
George Crump, CMO of Verge.io, described the situation this way: "Every byte sits on media that was purchased at 2026 prices. What hasn't been solved is the gap between the capacity companies buy and the capacity their workloads actually need." He further noted, "Reduce the number of architectures. Move cold data down a tier. Stop buying a second set of hardware just to protect the first set. None of this is exotic, but most companies have never actually measured any of it." With prices out of anyone's control, the very act of identifying wasteful, duplicate purchases has become a cost-reduction strategy in itself.
At the same time, the currently published commentary on the Omdia survey (commissioned by Verge.io) doesn't touch on technical considerations such as compatibility with existing hardware during migration to SDS, the cost of migration itself, or the impact on latency. There is still a considerable gap between 98% of North American IT professionals saying they're considering it and actual migration being completed.
Winners and Losers: NAND Makers' Revenue Growth vs. IT Departments' Cost Squeeze
The clearest winners in this situation are the NAND manufacturers. Kioxia, a Japanese company, is one of them. Its full-year financial results for the fiscal year ended March 2026, announced in May 2026, showed revenue of ¥2.3376 trillion (up 37% year-over-year) and operating profit of ¥870.3 billion (up 92.7% year-over-year)—the highest operating profit level in eight years. This marked the first time the company's revenue exceeded ¥2 trillion, with the joint production system with SanDisk at the Yokkaichi and Kitakami plants supporting a portion of the growth.
In January 2026, Kioxia and SanDisk announced they were extending their joint venture contract for the Yokkaichi plant by five years, from the end of 2029 to the end of 2034 (the Kitakami plant's contract already ran through the end of 2034 at that point). With both plants now set to continue operations through the end of 2034, the move signals a confident stance toward continuing business through the price surge.
By contrast, the ones bearing the brunt on the losing end are the enterprise IT departments buying the storage. Roughly three-quarters have been forced to slow down on-premises AI deployment, and they now have to rework their budgets under conditions where DRAM prices have nearly doubled and enterprise SSD contract prices have risen more than 50%. In the Omdia survey, 27% of respondents reported holding 1–10PB of data on-premises, roughly the same proportion reported 11–25PB, 18% reported 25–50PB, and 6% reported over 100PB. These four categories together account for roughly the high 70% range, with the remainder spread across other categories such as the 50–100PB range.The larger a company's scale, the more the price increases translate into hard dollar impact.
Additionally, more than a third of companies said 50–75% of their data was "dark data" of unknown utility, and 13% said the figure was 76–90%. The inefficiency of continuing to buy SSDs at current inflated prices for data whose very usage status is unclear has now become visible in concrete numbers.
The handling of dark data further complicates this picture. A situation in which more than half of a company's data holdings are of unclear utility means, in effect, that companies themselves cannot determine whether that data is a future asset for AI model training or simply a liability that they're continuing to store at inflated prices. Now that actual SSD prices have jumped several-fold, the cost of postponing that determination has grown far larger than before.
This shift in demand is also a business opportunity for SDS vendors. Companies such as VMware vSAN, Ceph, MinIO, Pure Storage, Weka, VAST Data, and VergeIO have found themselves fielding inquiries from companies looking to extend the life of existing hardware. While NAND manufacturers hold the reins on hardware pricing, a new competition is emerging over control of the software layer that sits on top of it.
How Outdated Is the 22.6x Figure?
The most eye-catching figure in this survey is the comparison Crump presented: "the cost per capacity of a 30TB QLC SSD is 22.6 times that of an HDD of the same capacity." This ratio, sharply up from 4.9x as of Q2 2025, is based on QLC SSD data as of Q1 2026 published by VDURA in April 2026. However, citing this peak figure as-is as "the latest gap" isn't quite accurate. Tracking what happened afterward reveals a trend that's more nuanced than simple continued widening.
Looking at actual 30TB TLC SSD and HDD prices tracked over time by StorageReview, the SSD-to-HDD price ratio surged from 7.0x in Q3 2025 to 23.2x in Q1 2026, before narrowing to 16.3x in Q2 2026, and standing at 18.6x as of the most recent Q3 2026 data. SSD prices themselves have continued rising consistently, from $3,460 in Q3 2025 to $22,600 (approximately ¥3.59 million) in Q3 2026—but when measured as a ratio, Q1 turned out to be the peak.
The reason the ratio narrowed isn't that SSD price increases let up. It's that HDD prices themselves surged over the same period, rising 61% from $755 to $1,216. Even HDDs—long regarded as the "cheap fallback" for data—have spiked sharply, compressing the relative gap with SSDs.
This fact bears directly on the very premise behind SDS's promised savings. The basic idea behind SDS is to combine expensive SSDs with cheap HDDs, sorting hot data from cold data to keep costs down. But if the per-unit price of HDDs themselves has more than doubled in a single year, the savings SDS promises may turn out to be smaller than initially expected. Using the 22.6x figure as a "still-valid yardstick" for procurement planning risks focusing solely on the SSD-side price surge while overlooking the simultaneous price increases happening on the HDD side.
Ripple Effects in Japan and Kioxia's Presence
Japanese companies are caught up in the same rising procurement costs as their North American counterparts. The enterprise SSD industry's Q1 2026 revenue of $18.5 billion translates to roughly ¥2.9415 trillion (at ¥159 to the dollar, as of August 13, 2026) for a single quarter alone, and Kioxia—one of the companies driving that figure—is a Japanese company with major plants in Yokkaichi, Mie Prefecture, and Kitakami, Iwate Prefecture. Because data center SSD and DRAM prices are determined by global supply and demand, companies in Japan planning to build on-premises AI infrastructure domestically cannot escape the same rising procurement costs.
Kioxia and SanDisk's decision to extend their joint venture through 2034 signals that they don't view the current price surge as a temporary supply-demand disruption. It reflects the NAND manufacturers' own judgment that this is a structural condition likely to persist for at least several years—and that rational business decision is, in turn, becoming a constraint that binds enterprise IT departments. This asymmetry won't resolve itself unless the semiconductor pricing cycle changes, and the growing interest in SDS is likely to become a fixture of procurement strategy built around the assumption of sustained high prices, rather than a passing trend.
The burden facing Japanese companies looking to refresh their server storage also comes into focus through concrete numbers. Now that a 30TB enterprise SSD has reached a price point of roughly ¥3.59 million per unit, procuring even a few dozen units alone moves a budget in the hundreds of millions of yen. Even setting aside currency fluctuations, this represents an order of magnitude difference from similar procurement plans just a few years ago. Given that ramping up new facilities inevitably involves a phased expansion of production volume, it's hard to imagine supply-side capacity increasing sharply in a short period.
With few levers available to bring prices down directly, the option left to companies is to change how they use storage. Tiered management and architectural consolidation through SDS serve as a stopgap measure for holding off on buying pricier SSDs, while also presenting an opportunity to surface wasteful, duplicate investments that companies had never even measured before. What lies behind the fact that 98% of North American survey respondents are taking action isn't so much a reaction to near-term price hikes as it is a structural sense of powerlessness—an inability to move hardware prices themselves. Until the production volume from Samsung and SK hynix's planned new investments ramps up in stages and is actually reflected in the supply capacity of NAND manufacturers, there's no prospect of this structure loosening.