Google and Microsoft are stepping on the gas: they prioritize speed over cost as hyperscalers compete for every gigabyte of memory

The new cloud gold rush is no longer GPUs, but memory. Several industry reports point to a growing mismatch between supply and demand for DRAM and HBM for data centers, and the major cloud service providers (CSPs) are reacting in very different ways. Among the US companies, Google and Microsoft are clearly prioritizing speed and availability over price, even accepting additional costs to secure preferential supply.

Memory Shortage Amid the AI ​​Race

The trigger is well-known: the explosion of generative AI and large-scale models has driven up demand for high-performance memory, both HBM (High Bandwidth Memory) for GPUs and DDR5 for servers. Adding to this, major manufacturers—Samsung, SK hynix, and Micron—are redirecting factories and production lines toward these premium products, reducing the availability of cheaper, traditional memory.

The result is a strained market where:

– Shipments to servers are prioritized over PCs and consumer devices.

– Some players have been aggressively stockpiling modules for months to ensure future capacity.

– DDR5 and HBM prices have seen double-digit increases throughout 2025, with particularly noticeable rises in high-capacity server modules.

Google and Microsoft: Paying More for Faster Development

According to memory channel sources and industry analysts, two clear strategies are emerging among US hyperscalers. On the one hand, Google and Microsoft are reportedly accepting higher prices to guarantee faster delivery and assured volumes for their AI clusters.

This makes sense:

– Both are engaged in a public race to deploy AI superclusters, Copilot-type services, Gemini, proprietary models, and enterprise AI solutions.

– A memory bottleneck would leave high-end GPUs underutilized, effectively wasting billions of dollars in hardware investments.

– For them, the extra memory cost is marginal compared to the strategic value of continuing to launch AI products at full speed and securing large enterprise contracts.

In other words, for Google and Microsoft, the «real risk» is not paying 20-30% more per module, but running out of RAM or HBM when their customers want to train models, fine-tune LLMs, or deploy AI agents on a global scale.

What does this mean for the cloud and AI market?

In the short term, the message is clear:

– Customers will see higher prices for GPU instances and high-memory nodes.

– Startups and smaller companies may find it more difficult to access cutting-edge AI resources at competitive prices.

– The gap between the large hyperscalers and the rest is widening, not only in software and models, but also in basic physical infrastructure.

In the medium term, this tension could accelerate:

– The development of more memory-efficient alternatives, such as architectures with better DDR utilization, new cache hierarchies, or more compact models.

– Increased interest in specialized on-premises and bare-metal solutions, where some companies prefer to invest in their own hardware to reduce their reliance on the cloud amidst escalating prices.

– Even greater regulatory and geopolitical activity surrounding chips, manufacturing, and export controls, with the US, Europe, Korea, and Taiwan as key players.

Source: www.revistacloud.com