KUALA LUMPUR, 21 March 2026 – The explosive growth of artificial intelligence is now threatening to create a new bottleneck in the global semiconductor supply chain, this time in data storage chips, as industry executives warn that demand could soon outstrip supply.
A senior executive from Solidigm, the US-based storage unit of South Korea’s SK Group, cautioned that AI’s rapidly expanding data requirements are set to strain storage chip availability for years, echoing earlier shortages seen in memory chips.
AI’s Data Appetite Is Growing Faster Than Expected
At the core of the issue is the sheer volume of data required to train and run modern AI systems. Unlike earlier computing workloads, AI models process vast datasets, requiring not just powerful processors, but massive storage capacity.
Industry estimates suggest that next-generation AI systems expected later this year could require around 35% more storage capacity than current models.
This surge is being driven by:
- Larger AI models and datasets
- Rapid growth in AI inference (real-time usage)
- Expansion of hyperscale data centres globally
As AI adoption accelerates across industries, storage is emerging as a critical but underappreciated constraint.
From Memory Shortage to Storage Crunch
The warning mirrors ongoing shortages in high-bandwidth memory (HBM), a key component used alongside AI processors, which are already under severe supply pressure.
The global semiconductor industry has been grappling with a structural shift, where manufacturing capacity is increasingly prioritised for AI-related chips, leading to imbalances across the broader ecosystem.
Now, storage chips, including NAND flash used in solid-state drives (SSDs), are at risk of becoming the next choke point.
Supply May Struggle to Keep Up Until 2030
Despite efforts to scale production and develop higher-density storage technologies, industry leaders are increasingly sceptical about keeping pace with demand.
Solidigm executives indicated that supply constraints could persist through the end of the decade, as production capacity expansion lags the exponential growth in AI workloads.
The challenge is structural:
- Building new semiconductor capacity takes years
- AI demand is accelerating faster than expected
- Supply chains remain concentrated among a few major players
Data Centres at the Centre of the Storm
The surge in demand is closely tied to the rapid build-out of AI data centres, which are becoming the backbone of the digital economy.
Global spending on AI infrastructure is projected to reach hundreds of billions of dollars, with tech giants racing to deploy more computing power and storage capacity.
This has created a cascading effect:
- More AI models → more data
- More data → more storage
- More storage → higher demand for chips
Implications for Markets and Technology
The potential storage crunch carries significant implications:
1. Rising Costs Across the Tech Ecosystem
Storage and memory prices could increase, affecting everything from cloud services to consumer electronics.
2. Investment Boom in Data Infrastructure
Companies are likely to accelerate investments in storage technologies, including next-generation SSDs and data optimisation solutions.
3. Strategic Shift in Semiconductor Supply Chains
Storage chips may become as strategically important as GPUs and AI accelerators.
Investor Takeaway: The Next AI Bottleneck
The AI revolution is no longer just about processing power, it is about data infrastructure.
While GPUs and advanced processors have dominated headlines, the emerging constraint in storage highlights a critical shift:
the future of AI will be determined not just by compute, but by the ability to store and move data efficiently.
The Bottom Line
The warning from Solidigm underscores a new reality in the AI era:
data is growing faster than the infrastructure built to support it.
As AI continues to scale, storage chips are poised to become the next battleground in the global semiconductor race, with supply constraints potentially shaping the pace of innovation through 2030.
