Nvidia and AMD Are Both Raising GPU Prices in 2026

The GPU market has never been kind to patience, but what's happening in 2026 has a particular quality of bad timing. Both Nvidia and AMD are raising prices—not just on flagship hardware aimed at enthusiasts with money to burn, but across the stack, including the budget tiers that have historically been the last refuge of the practical PC gamer.

There's no single smoking gun here. Instead, what you're watching is a cascade: memory costs up, wholesale kit prices up, retail prices up. The floor is rising.

How We Got Here

The year started with a round of increases that Game Rant put at roughly 10% to 15% on memory pricing in January, affecting both Nvidia and AMD products. That was the opening move. A second adjustment followed in May, focused primarily on Nvidia's top-end hardware—the RTX 5090 reached $5,900 for some of the highest-tier models, according to Game Rant. That number is so divorced from mainstream gaming budgets that it almost functions as a different market entirely. But the logic that pushed it there didn't stay contained.

Tech-insider.org reported that within the same 10-day window, Nvidia and AMD both raised the wholesale price of GPU-and-memory "kits" sold to their board partners—the downstream manufacturers who actually put cards on shelves. When you raise the input cost for the people making the product, the output cost follows. That's not mysterious; that's arithmetic.

The underlying driver is a memory shortage that has, per Tech-insider's reporting, "spent all year rewriting PC hardware price tags." GDDR6 and GDDR7 costs have climbed sharply, and WCCFTech has documented the downstream effects on China's retail market specifically, where GPU prices have seen massive increases as both memory cost escalation and retail stockpiling accelerate the pressure. Shops hoarding inventory in anticipation of further increases is the kind of self-fulfilling cycle that makes a bad situation structurally worse.

AMD's Particular Calculation

The more revealing detail in all of this isn't that prices are going up—it's how AMD chose to move. According to PC Gamer, AMD reportedly finalized its own price adjustments only after watching Nvidia's moves, having initially held back due to inventory still needing to be cleared and the fact that Nvidia's card prices hadn't yet visibly increased at retail.

Read that again. AMD waited for Nvidia to go first.

This isn't a secret about how markets work—price leadership is a documented phenomenon—but seeing it stated this plainly in the reporting is clarifying. AMD's budget cards have long been positioned as the sensible alternative for price-conscious builders. The competitive pressure AMD historically exerted on Nvidia in the mid-range and budget segments was genuine, and it kept prices from floating as freely as they might have otherwise. That pressure is now being used differently: as a signal for when it's safe to raise your own prices without losing market share, rather than as a tool for capturing it.

Kotaku put the situation with appropriate wryness: can't decide between Nvidia and AMD? Don't worry, they're both raising prices. The humor lands because the underlying dynamic is genuinely grim for anyone who was hoping that AMD's presence in the market would serve as a brake on Nvidia's pricing latitude. Right now, it isn't.

What This Looks Like at Street Level

Digital Trends offers one of the more grounding data points in all this reporting. The AMD RX 9060 XT—a card that launched for under $300—is currently difficult to find for less than $450. XFX's version runs $490. That's a 50%-plus premium over launch price on a card that was supposed to sit in the accessible tier of the market. Digital Trends also notes that this card is priced just $20 below the RX 9060 XT variant with double the graphics memory, which raises the further question of whether the lower-spec card is even a rational purchase at current prices.

Gizmodo cites Taiwan-based hardware tipster BenchLife as the source for reports that Nvidia was proposing yet another round of increases—suggesting the current wave isn't necessarily the last one. The tipster framing matters here: these are not official announcements from either company. They're sourced from supply chain observers and regional market watchers, which means some uncertainty is baked in. The directional signal, though, has been consistent across multiple independent outlets.

The Budget Tier Problem

It's worth sitting with what it means for the budget tier specifically to be caught in this. High-end GPU pricing has always been aspirational, and the audience for a $5,900 graphics card is small enough that its price is largely academic for most builders. Budget GPUs are different. They're the cards that let someone build or upgrade a PC without gambling a month's rent on it. They're what make PC gaming viable for people who aren't prioritizing it over other expenses.

When memory shortages and wholesale kit price increases squeeze that tier, the effects aren't evenly distributed. They land hardest on the people with the least flexibility—the ones who were already calculating how long they could run their current hardware before the next upgrade became unavoidable.

The current situation also complicates the console-versus-PC calculus that perpetually circles gaming communities. A mid-range GPU that was $250 two years ago and is now $400-plus starts to look a lot less compelling compared to a fixed-price console that ships with known specs. That math will occur to people without anyone having to explain it.

What the Record Shows and Doesn't Show

To be clear about the limits of what's currently documented: the specific scale and timing of upcoming increases beyond what's already been reported remains murky. The 10%-to-15% January memory increase and the May RTX 5090 adjustment are the most concrete figures in the record. The reports of further wholesale kit price increases from both companies are attributed to supply chain sources rather than official statements. Board partners haven't uniformly passed all costs through to retail simultaneously, which means street prices vary and will continue to vary by retailer and region.

What's not murky is the trend. Every data point across these sources points in the same direction, and the memory shortage driving it hasn't resolved.

The Part That Doesn't Resolve Cleanly

There's something worth naming directly here, because it doesn't fit neatly into the market-dynamics frame that most GPU price coverage uses.

The competitive relationship between Nvidia and AMD in the budget and mid-range GPU space used to mean something specific to the people building PCs on constrained budgets. It meant that neither company could get too comfortable, because the other one was waiting to take their customers. That competitive pressure—imperfect, inconsistent, but real—was one of the structural features that made PC gaming's hardware tier genuinely accessible across income levels.

Watching AMD reportedly sit on its own price increases, waiting for Nvidia to move first before finalizing its own adjustments, isn't illegal. It isn't even surprising once you've watched enough hardware market cycles. But it marks a shift in what AMD's competitive positioning is actually for. If the budget tier's primary purpose is now to follow the leader rather than undercut it, then the people who've extended their hardware cycle another year waiting for prices to normalize are doing math on a market that no longer works the way they remember it working. That's not nostalgia for cheaper prices. That's an accurate read on what's changed.

---

By Sarah O'Brien, Retro Gaming & Preservation Correspondent, Buzzrag