SanDisk’s new customer contracts and long-term financial targets have put the storage-chip maker at the center of a rally in memory-related stocks, as the company outlined a model built around multi-year purchase commitments rather than short-cycle spot demand.
At its 2026 financial presentation, SanDisk said it is targeting annual revenue growth of 15% to 19% from fiscal 2028 through fiscal 2030, alongside a gross margin of roughly 80%. The company also disclosed that it has signed new business model agreements with eight customers, covering about half of its expected fiscal 2027 bit shipments and roughly two-thirds of fiscal 2028 bit shipments.
SanDisk shares closed nearly 14% higher following the update. Micron rose 4%, while SK Hynix gained 7%, reflecting a market response that extended beyond one company’s outlook to expectations for tighter, more structured memory supply agreements.
Contracts cover a growing share of expected shipments
SanDisk said its new business model agreements include committed purchase volumes, binding contractual frameworks, minimum financial protections and structured pricing mechanisms. The agreements appear designed to give the company greater visibility over demand and pricing as large data-center customers seek dependable access to flash-memory supply.
“Bit shipments” measure the amount of memory capacity sold rather than the number of physical drives or chips. Covering half of fiscal 2027 bit shipments under contract and two-thirds of fiscal 2028 shipments would reduce SanDisk’s exposure to abrupt changes in the spot market, where memory prices have historically swung sharply as supply and demand shifted.
The structure also gives customers a clearer path to secure capacity over several years. That can be especially valuable for companies building AI data centers, cloud storage systems and large computing clusters that cannot easily replace memory components once infrastructure plans are underway.
SanDisk’s forecast of about 80% gross margin places considerable weight on the company’s ability to maintain pricing discipline, control manufacturing costs and preserve the value of contracted supply. Gross margin measures the portion of revenue remaining after the direct costs of producing goods. In semiconductor memory, that metric is often heavily influenced by changes in average selling prices and factory utilization.
The agreements do not eliminate market risk. Customers may still adjust deployment plans, while competitors can add supply or pursue their own long-term deals. Yet the size of SanDisk’s contracted shipment coverage suggests the company expects a larger share of its business to be governed by negotiated commercial terms rather than quarterly price moves.
Older Nvidia hardware remains commercially active
Nvidia Chief Executive Jensen Huang separately said that a recently signed contract for the company’s A100 GPU runs through 2029. The A100 was introduced in 2020 and has since been succeeded by newer products, including the H100, H200 and Blackwell-generation systems.
The contract shows that older AI hardware can remain commercially useful for years after newer chips enter the market. A100 processors are widely used for machine-learning training, inference and scientific computing workloads, and long-term agreements can keep those systems deployed even as customers expand fleets with more recent hardware.
That dynamic could support demand for components surrounding older GPU systems, including memory, storage, networking equipment and server maintenance. It also complicates the assumption that each new generation of AI hardware immediately displaces the prior one. Large computing operators often run mixed fleets, balancing performance needs against available power, software compatibility and the cost of expanding capacity.
Two measures could help indicate whether demand for older equipment remains firm: renewal pricing for older cards and utilization rates. Renewal pricing would show whether customers continue to pay meaningful rates to extend access to previous-generation GPUs. Utilization rates would reveal whether those systems are actively being used or simply retained as reserve capacity.
If both remain strong, it would indicate that older infrastructure is still contributing to available AI computing supply. Weakening lease rates or lower utilization would point toward a faster shift to newer hardware or a slowdown in workloads that can run efficiently on earlier generations.
Memory costs could intensify pressure on hardware budgets
JPMorgan Global Research has forecast that random-access memory costs could rise by 400% from 2024 through the end of 2026, according to the supplied material. Such an increase would place additional pressure on server builders and data-center operators already competing for advanced GPUs, high-bandwidth memory and storage components.
The memory market includes several distinct products with different supply chains. SanDisk’s core business is tied to flash storage, while random-access memory is generally associated with DRAM products made by companies such as Micron, Samsung and SK Hynix. Even so, price pressure across one category can affect overall server budgets and procurement decisions across the hardware stack.
Large technology companies with long-term supply agreements may be better positioned to secure capacity than smaller operators relying on short-term purchases. That can favor companies able to commit capital early, negotiate volume guarantees and absorb higher component costs.
Crypto claims require a separate evidence base
The reported developments in memory and GPU supply do not, by themselves, establish a direct effect on public blockchain hash rate or cryptocurrency prices. Hash rate measures the computing power securing proof-of-work networks, and its movement depends heavily on mining-specific hardware, electricity costs, network difficulty and the economics of mining each asset.
AI-oriented GPUs, enterprise storage and conventional RAM are not interchangeable with the specialized ASIC machines that dominate Bitcoin mining. A reported decline of 100 exahashes per second in global hash rate would therefore require network-specific data and mining-sector evidence before linking it to memory shortages or SanDisk’s contracts.
The more immediate market implication lies in the hardware industry: longer-duration purchase commitments could give memory suppliers greater revenue visibility while making capacity planning more difficult for customers outside the largest data-center groups.
For more on positioning around multi-year growth and hardware cycles, explore our digital assets market insights now.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
