Kioxia and SanDisk announced on August 27, 2026, that they expect to invest roughly ¥5 trillion in Japan through 2032. Contingent on government support and market conditions, the companies plan to continuously strengthen production equipment, related infrastructure, and technology at their Yokkaichi and Kitakami plants. The same day, they also revealed plans to build a third manufacturing building, "K3," at Kitakami, targeting operation within fiscal 2029. The new figure—equivalent to 55.6% of the roughly ¥9 trillion invested domestically over the past 25 years—marks a shift toward long-term expansion planning aimed at meeting AI-driven storage demand.
However, the roughly ¥5 trillion figure does not represent committed equipment orders. The companies have not disclosed year-by-year spending, the cost split between the two firms, the amount of government subsidy, or the resulting increase in production capacity. Even the timing of equipment installation at K3 will depend on market conditions. To understand what lies behind this large headline number, it helps to separate two things: the joint-venture structure the two companies have maintained for over 25 years, and the technology investments aimed at increasing the memory capacity extracted from each wafer.
The ¥5 trillion will be deployed in stages through 2032
The roughly ¥5 trillion figure equals 55.6% of the roughly ¥9 trillion invested cumulatively over the past 25 years. The intent to concentrate major investment in a relatively short period is clear, but the two companies have not finalized spending all at once. The announcement states they "expect to invest," explicitly conditioning this on government support and stating that equipment buildout will be aligned with market trends.
The K3 building, announced the same day, reflects this conditional approach concretely. Kioxia will begin land preparation on the south side of the K2 building at the Kitakami plant, aiming for operation within fiscal 2029. Details including construction timing and investment in production equipment will be decided later. The plan is to prepare the land first, then determine construction and equipment installation timing based on demand.
At Kitakami, capacity expansion is already underway without waiting for K3. The K2 building began operations in September 2025, and following initial 8th-generation production, 10th-generation 3D flash memory production started in July 2026. As equipment installation and technology upgrades continue at existing buildings, K3 will expand physical capacity beyond that. This investment framework layers together initiatives on different timelines.
When comparing figures, it's also important to note differences in scope.
| Figure | Scope | What it doesn't tell us |
|---|---|---|
| ~¥5 trillion | Investment Kioxia and SanDisk expect to make in Japan through 2032 | Annual breakdown, cost split between the two companies, subsidy amount, capacity increase |
| ~¥470 billion/year | Kioxia's capital expenditure plan for the next 3 years, announced June 2026 | SanDisk's share, total through 2032 |
| Up to ~¥242.9 billion combined | Combined cap of two past government subsidies approved in 2022 and 2024 | New subsidy amount tied to the current ¥5 trillion plan |
The ~¥5 trillion figure and the ~¥470 billion annual figure differ in company scope, time period, and expense category. They cannot simply be added together to produce a total investment figure, nor can either be straightforwardly converted into an annual rate. This announcement establishes an upper bound on the scale of investment, but disclosure of the actual cash flow is still to come.
A 50.1/49.9 joint venture links investment burden to production scale
At the core of manufacturing are three joint-venture companies: Flash Partners, Flash Alliance, and Flash Forward. Kioxia holds a 50.1% stake in each, with the SanDisk side holding 49.9%. These joint ventures own or lease the manufacturing equipment, while Kioxia manufactures wafers at Yokkaichi and Kitakami. The finished wafers are allocated between the two companies, each of which finishes them into SSDs or embedded products to sell to customers.
This arrangement allows the two companies to share the same front-end processes and technology development while maintaining their own product strategies and sales networks. Jointly procuring expensive manufacturing equipment makes it easier to build up production volume than either company could achieve supporting a factory alone. At a 2025 briefing, Kioxia explained that 20% of factory capacity is allocated exclusively to itself, with the remaining 80% split evenly between the two companies. This works out to an overall capacity allocation of 60% for Kioxia and 40% for SanDisk.
The joint-venture term has also been extended in line with the major investment. In January 2026, the two companies extended the Yokkaichi contract term by five years, from the end of 2029 to the end of 2034, aligning it with the Kitakami contract's end date. In exchange for manufacturing services and continued supply, SanDisk will pay Kioxia $1.165 billion between 2026 and 2029. This secures a contract period covering equipment expansion through 2032 and continued joint production beyond that.
That said, joint operation also locks in mutual dependence. If judgments diverge on equipment investment, technology transitions, or demand forecasting, neither company can complete its supply plans alone. The 25-plus-year track record and the contract extension through 2034 are stabilizing factors, but since the cost-sharing method for the roughly ¥5 trillion has not been disclosed, it remains unclear how far investment risk will actually be shared.
10th-generation 332-layer NAND boosts memory capacity per wafer
The 10th-generation BiCS FLASH, whose production began at Kitakami's K2 building, is a 1Tb TLC product stacking 332 memory cell layers. According to figures published by Kioxia, the NAND interface speed reaches 4.8Gb/s. Combined with higher planar density, bit density has been increased by 59%. The companies say write power efficiency improved by 18% and read power efficiency by 30%.
Production capacity cannot be measured by wafer input alone. Higher bit density increases the memory capacity that can be extracted from the same area. However, factors such as chip yield, product mix, and equipment utilization rates actually determine supply volume, so a 59% increase in density does not directly translate into a 59% increase in factory shipment volume.
CBA, used since the 8th generation, is a technology that forms the memory cell array and the control CMOS circuitry on separate wafers before bonding them together afterward. Because the memory side and the circuit side can each be manufactured using processes suited to them, it becomes easier to combine vertical improvements—increasing the number of stacked layers—with planar improvements that pack circuits and cells more densely. Factory investment goes not only toward new building construction costs but also toward the equipment, cleanrooms, and infrastructure needed for generational transitions.
As of July 3, 2026, the 10th-generation product was still at the sample-shipment stage, and specifications may change for mass-production units. Even so, production of this generation has already begun at the K2 building. The payoff from the roughly ¥5 trillion investment should be measured less by the number of new buildings and more by the speed of transition to new generations and how quickly stable yields of good product can be achieved.
AI inference is pushing NAND demand, but market volatility hasn't disappeared
One tailwind for this investment is the growing volume of data handled by AI inference. Even as GPUs handle computation, storage devices are needed to hold training data, models, and intermediate computation results. As inference is delivered to large numbers of users, demand grows for fast, high-capacity enterprise SSDs to line data centers. Kioxia has set a mid- to long-term target of raising the share of revenue from data center and enterprise products to 60% or more.
Current supply-demand conditions are tight. According to TrendForce, the enterprise SSD market in the first quarter of 2026 saw revenue rise 86.1% quarter-on-quarter to over $18.46 billion, with contract prices up roughly 80%. Kioxia reportedly progressed with qualification and mass production of existing-generation products among North American customers, while SanDisk also increased shipments of high-capacity products. This tightness is the reason both companies are moving to lock in supply capacity through 2032 ahead of time.
However, rising prices simultaneously reflect both strong demand and supply shortage. NAND is a market-cyclical industry where oversupply drives prices down, making investment recovery difficult. Even as AI-related demand grows, fluctuations in PC, smartphone, and consumer storage demand remain. That is why the two companies say they will decide on K3's construction timing and equipment investment details based on market trends.
There is no guarantee that AI demand will continue at the same pace through 2032. Treating the current surge in enterprise SSDs as a permanent growth rate risks turning expanded production into oversupply. What the two companies have announced is an investment framework that can flex with demand—not a plan to spend the same amount every year.
Government support drives the investment, but the new subsidy amount remains unspecified
The Japanese government has already supported joint production between the two companies. In 2022, a subsidy of up to approximately ¥92.9 billion was approved for Yokkaichi; in 2024, a subsidy of up to ¥150 billion was approved for 8th- and 9th-generation production at both Yokkaichi and Kitakami. The latter also serves a policy goal of dispersing production across two regions to reduce supply risk in the event of a disaster.
The current policy scale is even larger. The government's "AI and Semiconductor Industry Foundation Strengthening Framework" calls for more than ¥10 trillion in public support over the seven years through fiscal 2030, aiming to spur more than ¥50 trillion in combined public-private investment over ten years. The plans of Kioxia and SanDisk align with this direction of expanding Japan's advanced semiconductor capacity.
That said, the combined cap of roughly ¥242.9 billion from the two past subsidies cannot simply be added as the support amount for this new plan. The August 27 announcement includes no name or amount for any new subsidy program. While K3 is also premised on government support, construction costs and equipment installation timing have not been disclosed. The conditions that will turn the ¥5 trillion figure into actual spending include how far public funds will cover buildings, equipment, and R&D, and how much continued production the government will require in return.
The credibility of this investment plan will be confirmed not by the headline total but, in order, by subsidy decisions, annual spending, equipment deliveries to K3, and increases in bit supply. If these are disclosed and new-generation products can be brought online in step with demand, the two Japanese production sites will become a long-term manufacturing network supporting storage supply for the AI era.