Prices quadrupling in a matter of months
We have long since stopped talking about a slight indexing or limited price adjustment. In recent expansion filings, we saw RAM modules multiply in price by a factor of four or even five over what they cost a few months earlier. Enterprise SSDs show a similar story.
The peculiarity is that these price levels are also unstable: suppliers today give quotations that are valid for only two to three weeks, sometimes even only a fortnight, precisely because they themselves have insufficient certainty of supply and cost price.
The dollar rate this time is not the main culprit
Unlike the corona period, the dollar exchange rate plays a much more limited role today. Of course, some hardware remains USD-priced, but in recent cases we see that price increases do not follow proportionately from exchange rate fluctuations. The crux of the problem lies elsewhere: with the availability of critical raw materials, with energycosts and with production capacity in the semiconductor sector.
The Strait of Hormuz: an indirect but real impact
Recent geopolitical developments have added a new uncertainty factor. The (partial) closure and disruption of the Strait of Hormuz does not directly affect chip factories in Europe, but it does have a demonstrable indirect impact on global semiconductor production.
Multiple independent analyses show that:
- A significant portion of global LNG and energy exports through this route run
- Large semiconductor manufacturing regions such as Taiwan and South Korea rely heavily on imported energy to run their factories 24/7
- Helium, an essential gas for chip production (refrigeration, lithography), comes largely as a byproduct of LNG from the Gulf region. Where Qatar alone supplies about a third of global helium production
Disruptions in this chain do not necessarily lead to immediate production stops, but they do increase the cost, risk and volatility of memory and SSD production. Multiple analyses warn that memory chips (DRAM and NAND) are especially susceptible to this.
Specifically, what can you do about this?
In this context, it is important not to automatically start from “replace is better.” An often underestimated but very rational alternative is:
- Continue to use existing servers and storage for longer, especially if they are still technically adequate
- Manufacturers today often offer the ability to extend support to 7 years or more, including firmware, parts replacement and support
- Limited, targeted expansions (e.g., additional memory or storage where strictly necessary) can significantly extend service life
- After normalization of the market, you can re-evaluate and only then do a complete hardware refresh
Conscious risk management, not procrastination
Delaying purchases sometimes sounds like passivity, but in today’s context it is the opposite. Those who make complete hardware upgrades today at the peak moment of a volatile market are taking on unnecessarily high organizational costs. Those who couple timing with a clear analysis of the infrastructure state and market evolution are engaging in active risk management.
This requires an accurate view of the state of your current infrastructure, of what can be contractually extended and of the critical thresholds at which you really need to intervene. This is exactly what we at Xylos help our customers with: translating technical choices into an economically sound approach, with an eye for continuity today and smart timing tomorrow.
In conclusion
My message is not a plea against innovation, but rather for thoughtful timing. The current price pressure on RAM and SSDs is the result of structural factors in the global semiconductor supply chain, not temporary market fads. By being smarter about lifetime, expansion and purchase timing, organizations can bridge this period without sacrificing continuity or performance.
Want to know how to maximize your current infrastructure in these market conditions? Our team is happy to review with you the smartest approach for your specific situation.
About the author
Frank Dierickx is Partner Alliance Manager at Xylos and follows the evolution of hardware ecosystems, vendor relationships and infrastructure projects on a daily basis. His expertise helps customers make technology choices that are technically sound and economically justified.
