Memory shortage to 2030? The PC buying rulebook may be changing
TOP STORIES has been warning about the memory squeeze for a long time. Unfortunately, the news is getting worse.
Micron now says memory and storage supply will be even tighter in 2027 and 2028 than in 2026. More than 75% of its 2027 production is already committed, while new capacity coming online in late 2028 will still take time to ramp.
And 2028 may NOT be the end.....
SK Hynix has gone further, suggesting the shortage could persist through the end of 2030.
That is not a certainty. Additional Chinese production, weaker demand or another semiconductor downturn could change things. But this is increasingly looking less like another short memory cycle and more like a structural change in the supply chain.
And that raises another interesting question..... who really makes your computer?
Names such as Tongfang, Clevo, Quanta, Compal, Wistron, Pegatron and Foxconn mean very little to most computer buyers, yet companies like these form much of the ODM and contract manufacturing ecosystem behind the global PC industry.
An ODM, Original Design Manufacturer, can design and manufacture an entire computer platform which is then customised, qualified and sold under somebody else's brand.
It means the famous logo on the lid does not necessarily identify the company ACTUALLY assembling it.
This has always been how much of the electronics industry works, including LCD screens..... but that is another story.
What may change is how closely customers examine the markup between the underlying hardware and the finished branded product.
Harder Questions
If component shortages continue towards 2030, buyers are going to ask harder questions..... what exactly am I paying the premium for?
There are VERY good answers, so DO NOT rush out and start buying unknown computers just yet.....
Dell, HP and Lenovo do far more than put a badge on an ODM computer.
Enterprise customers are paying for controlled configurations, qualified components, BIOS updates and firmware management, drivers, security, international warranties, spare parts, predictable product roadmaps, fleet management and support years after the machine was purchased.
For organisations managing thousands of endpoints, that has considerable value.
What's changing is that the value can no longer simply be assumed.
Apple has already spotted something happening at the other end of the market.
The MacBook Neo starts at just US$599, deliberately opening the Mac ecosystem to buyers who previously might never have considered Apple. Instead of forcing every buyer up into a MacBook Air, Apple created a lower-cost product for users who simply do not need everything the premium machine provides.
That is an important change in product selection behaviour..... buy what the user actually needs, rather than automatically buying the traditional product above it.
Others are already responding, with new Windows machines being positioned directly against Neo.
The same thinking could spread into corporate hardware procurement.......
- Does every employee require exactly the same premium business notebook?
- Does every device need to follow the same three or four-year replacement cycle?
Premium enterprise devices will remain important where support, security and lifecycle management justify the premium.
Other users may increasingly be served by longer refresh cycles, alternative form factors, refurbished equipment, simpler configurations or hardware much closer to the ODM end of the market.
That does NOT mean buying unknown computers because they are cheap.
It means understanding where the value really sits.
The computer industry may increasingly move from buying a brand to buying the appropriate level of lifecycle assurance for each user, with attention moving to how many resources the user is consuming from a licensing perspective....
Premium brands are not going away..... but increasingly, the premium will need to be proved, rather than simply accepted.
If the memory squeeze really does extend towards 2030, buying computers the way we always have may become one of the first assumptions that needs to change.
|