Framework RAM Price Hikes: The DIY PC Crisis Explained [2025]

If you've been paying attention to the PC market lately, you've probably noticed something weird happening. Prices are all over the map, especially for RAM and SSDs. But here's the thing that's really interesting: the pain isn't distributed equally. Big companies like Apple? They're barely feeling it. Smaller PC makers? They're getting crushed.

Framework, the company behind those modular, repairable laptops that everyone got excited about a few years ago, is right in the crosshairs. The company has been raising prices almost monthly since late 2025, and the increases aren't small. An 8GB RAM stick that cost

This isn't just a Framework problem. This is a warning sign about what happens when supply chain disruptions hit an industry that's already fragile. And it's getting worse before it gets better.

TL; DR

• Monthly price hikes: Framework has raised RAM and system prices every month since November 2025, with increases ranging from 6% to 16% as reported by PCMag.
• Brutal cost jumps: 8GB RAM modules went from
  $40 to$
  130 in six months; 96GB kits tripled from
  $480 to$
  1,340 according to Ars Technica.
• Desktop impact: Maxed-out Framework Desktop systems jumped
  $600 from launch pricing to$
  2,599.
• Why it matters: Smaller PC makers lack Apple's buying power, forcing them to pass full costs to consumers as noted by Marketplace.
• Future outlook: Framework expects prices to get worse before stabilizing, citing AI-driven memory demand.

Why RAM Prices Are Exploding Right Now

Let's start with the obvious question: why is RAM so expensive all of a sudden?

The short answer is AI. Everyone wants AI now. That means everyone's buying high-capacity memory modules, everything from data centers training massive models to companies upgrading their infrastructure to run local AI workloads. When demand spikes that dramatically, supply can't keep up. And when supply can't keep up, prices go up.

But that's the surface-level explanation. The real story is more nuanced.

Memory manufacturers like Samsung, SK Hynix, and Micron have been planning their capacity expansion for years. They can't just flip a switch and make more RAM overnight. Chip fabrication plants are expensive, complicated, and take time to build. So they planned their expansion based on what they thought demand would be. Then AI happened, and suddenly their projections looked laughably conservative as highlighted by South China Morning Post.

Now here's where it gets brutal for companies like Framework. The big players—Apple, Dell, Lenovo, HP—signed long-term contracts at fixed prices before the crunch hit. Apple's basically locked in. When supply got tight, they could absorb the difference or negotiate better terms because of their massive volume. Dell has similar leverage. Framework? They're buying memory on the spot market like everyone else, and spot prices have become absolutely unhinged as discussed in The New York Times.

Patel, Framework's CEO, was pretty transparent about this in a recent livestream. The company's essentially passing through its actual costs to customers because it doesn't have the negotiating power to absorb the hits. In September, Framework's 8GB SODIMM cost them roughly

The really wild part? Framework isn't even the worst case. Companies like Raspberry Pi and Valve have faced similar constraints. But Framework's situation is particularly painful because RAM is core to their value proposition. You buy a Framework laptop specifically because you can upgrade it. When the upgrade parts cost three times more than they did six months ago, suddenly that value proposition feels a lot weaker.

The Math Behind Framework's Price Increases

Let's look at some specific numbers, because they're actually pretty shocking when you line them up.

In September 2025, here's what Framework charged:

• 8GB SODIMM: $40
• 16GB SODIMM: Around $80 (estimated)
• Framework Laptop 13 with 16GB: Base configuration around $1,200
• Framework Desktop with 32GB: $1,099 at launch

Now, in early 2026:

• 8GB SODIMM: $130
• 16GB SODIMM: Around $260 (estimated, based on doubling pattern)
• Framework Laptop 13 with 16GB: Prices holding steady for now, but Framework expects increases
• Framework Desktop with 32GB: $1,209
• Framework Desktop with 128GB: $2,599

So the Desktop with maximum RAM went from some launch price to

Why the Desktop got hit harder than the Laptops makes sense when you understand the difference. The Laptop 13 uses user-replaceable SODIMM modules. The Desktop uses soldered LPDDR5X memory. Soldered memory is more efficient, gives better performance, but you can't upgrade it. When RAM gets expensive, you really feel it in systems where you can't swap it out later.

Framework's been honest that they're pricing these modules "as close as we can to the weighted average cost of our purchases from suppliers." That's a fancy way of saying: we're barely marking these up at all. We're just passing through what we're actually paying.

Here's a helpful formula for understanding memory cost impact:

For the 8GB module:

For the 96GB kit:

Those aren't percentage increases. Those are complete market disruptions.

Apple vs. Framework: Buying Power Tells the Story

Here's what's genuinely infuriating about this situation: it's not that memory got expensive for everyone equally. It got expensive, but the pain distribution is completely uneven.

Apple buys memory in mind-boggling quantities. When you're ordering tens of millions of memory modules per year, you have leverage. You can negotiate long-term contracts. You can lock in prices. You can even threaten to shift supply to competitors if a manufacturer doesn't play ball. Apple's basically at the table as an equal with Samsung and SK Hynix. They can argue about price.

Framework orders memory in the hundreds of thousands. Maybe. Compared to Apple, that's rounding error. Framework doesn't get a seat at that negotiating table. Framework buys from distributors, who buy from manufacturers, who are dealing with their own allocation problems.

When spot prices for DDR5 modules skyrocket, Apple feels it as a rounding error in their quarterly earnings. Maybe their Mac Book Pro margin drops from 42% to 41%. For a company making $400 billion in revenue, that's a non-event. They just engineer around it or absorb it.

For Framework, a

This is why Framework wrote that statement basically saying: "We don't have the buying power of Apple, Dell, Lenovo, or HP, so we're just going to be transparent about what's happening and pass costs through as fairly as we can." They're not being greedy. They're being realistic about leverage they don't have.

The same calculus applies to Valve with Steam Deck production, or Raspberry Pi with their single-board computers. None of these companies have the scale to negotiate with memory manufacturers the way the big OEMs do.

Monthly Price Hike Patterns and What They Reveal

Framework stopped selling standalone RAM in November 2025. That's a pretty significant decision that most people glossed over. Why'd they do that?

Probably because the situation got so volatile that they couldn't keep pricing stable enough to make sense. If you price RAM today and memory costs jump 15% in the next week, you're either eating losses or you're constantly updating prices. At some point, it's cleaner to just stop selling the standalone component and bundle it into system pricing.

Since then, Framework's raised prices on systems every single month. Sometimes it's the Desktop. Sometimes it's the Laptop DIY editions. Sometimes it's SSD bundles. But the pattern is consistent: something goes up almost every month.

Here's the timeline they've been dealing with:

• November 2025: Stopped selling standalone RAM, began monthly system price adjustments
• December 2025: First round of price hikes on Desktop and Laptop configurations
• January 2026: Continued increases, particularly on high-capacity configurations
• February 2026: This week's announcement of additional increases hitting Desktop and DIY Editions hard

The increases are scaling with RAM quantity. Low-capacity systems saw 6% increases. High-capacity systems saw 16% increases. That tracks perfectly with spot market memory pricing.

Framework CEO Nirav Patel was pretty candid in a livestreamed Q&A: they're trying to fix problems everywhere they can, but sometimes the best they can do is be transparent about what's actually happening. That's a genuine acknowledgment that they're not choosing to raise prices. They're being forced to.

Why DIY and Upgradeable PCs Got Hit Hardest

Here's an interesting market dynamic that most people haven't thought about: the DIY and upgradeable PC market is actually more vulnerable to memory price spikes than the locked-down, proprietary market.

Why? Because with DIY and upgradeable systems, customers expect memory to be an aftermarket upgrade. They buy the base system and add more RAM later. When RAM becomes expensive, customers don't buy upgrades. They buy less than they originally planned.

With locked-down systems like Apple's, you buy it with the memory you want at purchase time or you're stuck. If you want 64GB, you have to commit at purchase. With Framework, you're "supposed" to be able to buy 32GB and upgrade to 64GB later when prices hopefully come down.

Except prices aren't coming down. They're going up.

Framework tried to position the DIY Edition as a partial solution. Want to buy a Framework Laptop without RAM and source cheaper modules elsewhere? Go for it. You can get a RAM-less Laptop 13 or Laptop 16 and populate it with your own modules from other sources.

But here's the thing: if memory's expensive at Framework, it's probably expensive everywhere. The spot market is the spot market. You might find someone liquidating old inventory somewhere, but you're not going to find a wholesale deal that beats what Framework could negotiate.

Framework also did a small giveaway for spare RAM sticks from "the bin of random modules in our office." That's honestly kind of touching in a sad way. They're literally giving away random old modules because they have them and customers need them. Twenty people winning a stick doesn't solve the RAM crisis, but it shows the company's aware of the pain they're causing.

The SSD Situation: Another Pricing Ticking Time Bomb

Let's talk about SSDs, because that's the next shoe about to drop.

Framework's SSD pricing has been flat so far. But Patel explicitly said they're selling 8TB SSDs "for substantially below the available market pricing." That's a pretty clear signal that they're eating the difference right now and prices will go up soon.

SSD pricing has followed a similar pattern to RAM, but with a twist: higher-capacity drives got hit harder than lower-capacity ones. A 1TB SSD didn't increase much. An 8TB SSD? Prices absolutely exploded.

This makes sense from a manufacturing perspective. NAND flash capacity expansion is expensive, and when demand is high, manufacturers prioritize high-margin, high-capacity products. The infrastructure to make 8TB drives is more capital-intensive, so when demand outpaces supply, those get priced higher.

Framework knows this is coming. They're being transparent about it. They're essentially warning customers: buy your SSD upgrade now, or expect prices to jump.

Memory Manufacturers' Capacity Constraints and Timeline

So when's this getting better? That's the million-dollar question everyone's asking.

Memory manufacturers have been announcing capacity expansions for years. Samsung's building new fabs. SK Hynix has expansion plans. Micron's investing in new production. But here's the reality: building a semiconductor fab takes 3-5 years and costs billions of dollars. You can't just flip a switch.

Samsung announced some new capacity coming online in 2026 and ramping through 2027. SK Hynix has similar timelines. But "coming online" doesn't mean "cheap immediately." When new capacity first comes up, manufacturers typically run at capacity-building-but-also-maintaining-high-margins. Prices don't typically drop immediately according to Deloitte's semiconductor industry outlook.

Industry analysts are projecting that memory prices might stabilize sometime in late 2026 or early 2027. But "stabilize" might just mean "stop going up quite as fast." Don't expect prices to crash back to 2024 levels.

Framework's statement that things will "get worse before they get better" is probably accurate. They might see prices continue climbing through Q2 2026, then stabilization in H2, then maybe modest decreases in 2027.

That's a 12-month horizon of pain for customers trying to buy or upgrade their systems.

The Broader PC Market Impact Beyond Framework

Framework's situation is actually just a window into a much bigger market problem.

Valve reported similar struggles with Steam Deck pricing. Raspberry Pi had to discontinue products because they couldn't source components at profitable prices. Even big players like Dell and Lenovo have had to adjust pricing, though with much more sophisticated strategies to minimize customer impact.

The reason Framework's story got attention is because they're a smaller, more transparent company. They don't have a PR department that smooths over price increases with marketing language. CEO Nirav Patel basically just explains: "Hey, memory costs went up, so our prices are going up. We don't like it either."

That honesty is refreshing in an industry that usually hides price increases behind "supply chain optimization" and "market adjustments." But it also shows how fragile smaller manufacturers are when component costs spike.

Longer term, this might actually change PC buying behavior. If upgradeability becomes more expensive because upgrade components cost so much, more people might just buy fully-configured systems upfront. That means fewer Framework Laptops sold with base configurations and upgrades added later. That means more sealed systems.

Which is ironic because Framework's whole pitch is that sealed, unrepairable systems are bad. But when the upgrade components cost three times more than they did a year ago, that pitch loses some power.

Why This Matters: The Death of Repairability Economics

Here's something that doesn't get discussed enough: expensive upgrade components can actually kill the repairability market.

Framework's value proposition depends on upgradeable hardware. You buy a Laptop 13 with 16GB, keep it for three years, then add 32GB more. You replace the SSD when you need more storage. You swap out the display if yours gets damaged.

But if RAM costs $130 for an 8GB module, suddenly buying a brand new laptop with 32GB is cheaper than upgrading your existing one. A used Framework Laptop with 16GB of soldered RAM becomes less valuable because you can't upgrade it cost-effectively.

This creates a weird economic situation where repairability—the entire reason Framework exists—becomes less attractive than buying new hardware. That's backwards from what the company intended.

There's also the environmental angle. Framework's whole pitch includes a sustainability message: keep your hardware longer by upgrading components instead of replacing the whole system. When upgrade components become prohibitively expensive, you've effectively eliminated that environmental benefit.

Nirav Patel's probably thought about this. It's probably keeping him up at night. Because on some level, he's created a company whose core business model depends on component prices remaining relatively stable. A 200% spike in memory pricing breaks that model.

The DIY Edition might be a temporary fix. Customers can buy the base hardware and source cheaper RAM from secondary markets, liquidation sales, corporate surplus vendors. But that's not a sustainable long-term strategy. That's a workaround.

The AI Demand Explosion and Memory Economics

Let's talk about why this is happening in the first place: AI demand is completely insane.

Every company wants to run AI. Every device wants AI features. Every developer wants to build AI products. Data centers are buying memory by the truckload. Consumer devices are increasing default RAM just to handle AI workloads. Edge devices for local AI inference need more capacity than they used to.

The memory market went from "normal demand with predictable growth" to "holy crap, everyone wants memory right now" in about six months.

And memory manufacturers, despite having expansion plans, can't scale fast enough. They're ramping capacity, but the ramp isn't as steep as demand is climbing.

Here's the physics of it: semiconductor fabs are the most capital-intensive manufacturing facilities in the world. A modern memory fab costs $5-10 billion to build. It takes 3-5 years to build and commission. By the time it's online, you've invested a decade of planning, capital, and risk.

You can't just decide to make more memory faster. You plan capacity expansion years in advance based on demand projections. Then AI happened and rendered those projections obsolete.

So now the industry is in a weird situation where everyone's expanding capacity as fast as they can, but "fast" in semiconductor terms means 2-3 years. Meanwhile, demand is growing month to month.

This creates a supply crunch that favors big customers with negotiating power and hurts smaller customers without it. Which is exactly what we're seeing with Framework.

Framework's Transparency Strategy and What It Reveals

Most companies hide price increases. They rebrand products so it looks like you're getting more features when really you're just paying more. They phase out old models quietly and move inventory to slightly different SKUs at higher prices.

Framework's done almost the opposite. They published a blog post explaining their costs. They did a livestreamed Q&A where Patel addressed the situation directly. They admitted they don't have the leverage to fight this. They explained their margins.

That transparency is either brilliant or terrible depending on how you look at it.

Brilliant because it builds trust. Customers understand that Framework isn't gouging them; they're getting squeezed by the same market forces squeezing everyone else. That honesty creates loyalty.

Terrible because it makes the problem visible. If Framework had just raised prices quietly and hoped nobody noticed, maybe some customers wouldn't have cared. Instead, by being transparent, they've drawn attention to exactly how much component costs have spiraled.

But I think the transparency was the right call. Customers would have figured it out eventually. Better to explain it clearly and maintain trust than get caught price-gouging and lose credibility.

Also, there's something kind of gutsy about a CEO just saying "we're not in control of this situation." Most corporate leadership spends energy talking about what they've accomplished or what they're planning. Patel's saying: "We're dealing with market forces we can't influence. Here's what we're doing about it."

That's refreshingly honest.

Solutions Framework Is Trying (And Why They're Partial Fixes)

Framework's not just complaining. They're trying to offer solutions, even if they're imperfect.

DIY Edition without RAM: Buy the laptop without RAM pre-installed, source your own modules. Problem: there's no guarantee you'll find cheaper modules elsewhere. If memory's expensive on the spot market, it's expensive everywhere.

The RAM bin giveaway: Give away spare modules to lucky winners. Problem: this helps 20 people. Framework probably sells thousands of laptops. It's a nice gesture that barely dents the actual problem.

Transparency about SSD pricing: Warn customers that SSD prices will go up soon. Problem: this doesn't prevent the price increase, it just gives customers time to buy before it happens. That's helpful but not a solution.

Focus on modularity and upgradeable design: This is the real long-term answer. Customers buy a base system, then upgrade components when memory prices stabilize. Problem: this requires memory prices to stabilize, which won't happen for another 6-12 months at least.

Framework's basically doing what they can within the constraints they're facing. They can't manufacture their own memory. They can't negotiate directly with Samsung or SK Hynix. They can't change spot market pricing. So they're passing costs through and being transparent about why.

Competitive Positioning: How Other PC Makers Are Handling This

Dell, Lenovo, and HP have handled this more gracefully because they have options Framework doesn't.

First, scale. When you're Dell and you buy hundreds of millions of memory modules per year, you get better pricing and better terms. You can negotiate multi-year contracts at fixed prices. When spot market prices spike, you're protected by existing contracts for a while.

Second, product variety. Dell makes everything from cheap laptops to high-end workstations. If high-capacity RAM gets expensive, they just don't offer high-capacity configs on budget models. Customers don't expect a

Framework's core products are premium. Their Laptop 13 starts around $1,200-1,300. Customers expect premium specs. When those specs get expensive, there's no room to pull back without making the product feel cheap.

Third, brand elasticity. Apple raised Mac Book prices slightly and nobody really cared because Mac Books always cost money. Framework's building a brand based partly on "repairable, upgradeable, fair pricing." When prices spike 200%, it feels like a betrayal of the brand promise.

Apple's also willing to eat costs. They can afford to have margin compression. Framework probably can't. Framework's trying to stay profitable while passing through actual costs.

So Framework's situation isn't unique—all PC makers are dealing with expensive memory. But Framework's transparency, smaller scale, and brand positioning have made it more visible.

What Customers Should Do Right Now

If you're thinking about buying a Framework system, here's my honest take:

If you want the laptop, buy it now. Prices are going up every month. By next month, it'll cost more. By next quarter, it might cost significantly more.

Buy it with the configuration you actually need. Don't plan on upgrading RAM later to save money, because RAM won't be cheaper later. It might be the same price or more expensive.

The DIY Edition might make sense if you've got access to cheap used RAM from somewhere (corporate liquidation, eBay marketplace deals, secondhand computer resellers). But don't plan on this. Most people won't find better pricing elsewhere.

For SSDs, same logic applies. Buy before prices increase if you can. Framework's warned that SSD pricing will rise soon. When it does, it'll probably be significant.

Longer term, this situation probably favors sealed, pre-configured systems over upgradeable ones. If you're planning to keep a laptop for 5+ years and upgrade along the way, a Framework makes sense. You get repairability and you're not locked into whatever configuration you bought.

But if you just want a good laptop and don't think you'll upgrade components, a Dell or Lenovo with fixed configuration might be better. You'll avoid paying premium prices for upgradeable hardware you won't use.

The Broader Industry Lesson: Fragility of Global Supply Chains

Framework's crisis reveals something uncomfortable about how tech supply chains actually work.

We talk a lot about "global supply chains" like they're these robust, efficient systems. In reality, they're fragile equilibriums that work as long as nothing unexpected happens. The moment demand or supply patterns change unexpectedly, smaller players get squeezed.

Framework has limited options because they're small. Big companies have options: absorb costs, negotiate better terms, shift products, raise prices with less resistance. Framework has basically one option: pass costs through and hope customers stick with them.

This has implications for the broader push toward repairability and sustainability. If repair-focused companies can't compete economically during supply chain stress, they're going to struggle to grow. And if they can't grow, the whole repair-focused PC market remains niche.

Apple's sealed iPhone? That's vulnerable to the same supply chain stress. But Apple's scale gives them resilience that Framework doesn't have. When component costs spike, Apple can afford to eat it or make strategic product changes. Framework has to pass it through because they don't have cushion.

There's probably a lesson here about how tech markets naturally consolidate around big players, because big players are better insulated from supply shocks. Framework's not going out of business or anything, but they're being tested harder than Dell or Lenovo.

Future Outlook: When Does This End?

Framework's statement that things will "get worse before they get better" is probably accurate. Here's why:

Memory demand is still growing. Every company launching a new AI device needs more capacity. GPT adoption is accelerating. Local AI inference is becoming more common. Consumer devices are adding more RAM by default.

Meanwhile, memory manufacturers are ramping capacity, but not fast enough to meet demand growth. The ramp is happening, but the demand growth is faster.

So for the next 6-12 months, we're probably in the painful part of the cycle. More supply is coming online, but not fast enough to relieve pressure. Prices probably keep rising or stay elevated.

Sometime in late 2026 or early 2027, capacity additions start catching up to demand growth. That's when prices start stabilizing. But stabilizing doesn't mean dropping back to 2024 prices. It probably means high 2025/early 2026 prices become the new normal.

Forge a bit further out: by 2028-2029, memory capacity should be abundant again. Prices might actually drop from the current insane levels. But we're talking 2+ years away for that.

Framework has to survive the next 12-24 months at elevated component costs with their current business model. That's the challenge they're facing.

Conclusion: A Stress Test on the Whole System

Framework's price hikes aren't really about Framework. They're a window into how market disruptions actually affect companies and consumers.

Big companies have leverage. Small companies don't. When component costs spike, that difference becomes painful and visible.

Framework's handling it about as well as they can: transparently, honestly, and without pretending they have control they don't have. That's probably the best strategy available to them.

For customers, the lesson is clear: if you want repairability and upgradeability, commit now while prices are elevated but not at their worst. If you want to wait for stable pricing, you're looking at mid-to-late 2026.

For the broader industry, the lesson is uncomfortable: smaller, more sustainable, more repairable companies are more vulnerable to supply chain stress than consolidated, vertically integrated giants. That probably favors consolidation and sealed systems unless the industry finds ways to build resilience for smaller players.

Memory prices will stabilize eventually. Supply will catch up to demand. But the path to get there is going to be painful for customers and stressful for companies like Framework that don't have Apple's scale or pricing power.

The DIY PC market isn't dead. But it's been tested, and the test shows that it's more fragile than it seemed six months ago.