There’s a lot to like about Framework, from its modular approach to gaming laptop upgrades and hardware repairability, to the company’s candour regarding the memory supply crisis.
Pre-orders for a 12 GB version of the Framework Laptop 16 graphics module recently went live. Powered by the Nvidia GeForce RTX 5070, it’s fair to expect this to go for a pretty penny—but at $1,199, the 12 GB pre-orders cost twice as much as the 8 GB modules.
TechPowerUp highlighted this fact on X, to which the official Framework account responded, “Thanks @OpenAI.” When another user argued that the 12 GB module should cost less, Framework replied, “Feel free to start a laptop company and see what suppliers quote you for GDDR7.”
The ongoing memory shortage is at least in no small part fuelled by the AI industry’s nigh insatiable appetite for SSDs and RAM. Framework itself had warned that stabilising memory prices were only a ‘temporary reprieve’ earlier this year. Indeed, prices are likely to look a lot worse before they get better; last month company CEO Nirav Patel even said, “If you’re looking to order a Framework Laptop or Desktop with a lot of storage, now is the time to do that.”
What’s fuelling the fire of the RAMpocalypse is outsized demand. As such, memory manufacturers are trying to scale up—but the new factories being constructed by major players such as SK Hynix, Micron, and Samsung won’t make a meaningful impact on supply until about 2028.
Feel free to start a laptop company and see what suppliers quote you for GDDR7.April 29, 2026
With such supply challenges in mind, $1,199 for a 12 GB graphics module still warrants a sharp in-take of breath, but also seems a touch more understandable.
As for GDDR7 and Nvidia, some leaks suggest we’ll see a whole 9 GB of it in a release of the RTX 5050 GPU later this year. If that turns out to be true, this card will use three sets of 3 GB chips rather than four sets of 2 GB. That could help with cost, in theory—but I dread to think just how much this entry-tier card could end up asking for in practice.
