Enterprises grappling with surging memory prices and constrained supply are increasingly turning to cloud infrastructure over on-premises deployments. Amazon CEO Andy Jassy flagged the trend during the company’s latest quarterly earnings call.
Jassy stated the cost of key components, particularly memory, has skyrocketed, while supply has failed to keep pace with demand. This imbalance is creating new pressures for organizations and is pushing companies that have on-premises infrastructure into the cloud.
“We have seen a number of conversations we have been having with enterprises for many months, where it has just been slower in getting the transformation plan to move to the cloud accelerate rapidly just because we have a lot more supply than what others have. It will be interesting to see how that evolves over time,” Jassy said.
The trend is reflected in the strong revenue growth reported by hyperscalers, signalling sustained enterprise demand for cloud infrastructure.
“Q1 FY2026 numbers of hyperscalers support stronger cloud demand. Google Cloud grew 63% year-over-year in Q1 2026, AWS posted 28% growth, and Microsoft Azure hit 40% growth. The cloud growth rate has been highest in recent quarters,” said Pareekh Jain, CEO at EIIRTrend & Pareekh Consulting.
However, Jain cautioned that this surge is not entirely attributable to memory constraints. The acceleration is also being driven by rising enterprise investments in AI workloads, making it difficult to conclusively link the growth solely to shortages in high-bandwidth memory (HBM).
Supply constraints reshape infrastructure decisions
The comments come amid a broader supply crunch driven by AI-led demand for HBM and advanced DRAM. Leading memory giants are realigning businesses to meet the rising demand.
For instance, Micron Technology has scaled back its Crucial consumer business globally to redirect output and investment toward enterprise-grade DRAM and SSD products.
As a result, cloud providers, including AWS, have moved early to secure large volumes of these components through long-term supplier agreements, giving them a clear advantage over enterprises procuring hardware.
Jassy acknowledged identifying the trend in the middle to latter part of last year, following which the company has been working closely with its strategic partners and suppliers to secure a significant amount of supply. As a result, currently, AWS is not capacity-constrained but the company will continue to watch the developments closely.
Cloud TCO shifts
This dynamic is beginning to reshape CIO decision-making. Rather than committing capital to increasingly expensive and difficult-to-source infrastructure, enterprises are turning to cloud deployments that shift spending to an operational model while ensuring access to scarce resources.
“We will certainly see an increase in public cloud adoption this year and next as cloud prices have either not increased, or have marginally gone up. Accessibility for cloud has improved because of availability, immediate availability vs long lead times for server procurement, and a different cost trade-off. When an on-premises server costs 4x what it did a year ago that changes the comparison to cloud, and that’s what is driving near-term TCO decisions,” said Shrish Pant, director analyst at Gartner.
Jain added that, in terms of pure unit cost, on-prem remains 40–50% cheaper for steady-state workloads if you can actually get the hardware. However, the TCO of waiting 9 months for a server includes lost revenue and missed market opportunities, which makes the cloud’s higher rent attractive to CFOs.
But for mission-critical workloads, enterprises are likely to retain on-prem deployments where capacity can be secured, reinforcing a hybrid approach rather than a full shift.
Vendor lock-in shifts to the supply chain
Given the current environment, where large cloud providers are securing priority access and even supplier attention, enterprises procuring smaller volumes are increasingly facing higher prices, longer lead times, and weaker negotiating leverage. As cloud becomes the only environment where capacity can be provisioned quickly, workload placement decisions risk becoming less strategic and more driven by necessity, warn analysts.
“The newer dependence is not primarily about whether software runs on someone else’s infrastructure. It is about whether equivalent compute capacity, with equivalent power, on equivalent timelines, is even procurable for an enterprise of average size and average leverage,” said Sanchit Vir Gogia, chief analyst and CEO at Greyhound Research.
He cautioned that vendor lock-in now begins before the workload is migrated. So CIOs who fail to account for this shift risk losing negotiating leverage. The CIOs who understand that will negotiate accordingly. The ones who do not will discover, a contract cycle from now, that the optionality they assumed they had was quietly handed over the moment they signed.
