When Great GPUs Meet Terrible Pricing
2025 should have been an amazing year for PC gamers. Both AMD and Nvidia dropped big new GPU generations. On paper we got serious performance, new architectures and cards that promised high end gaming at more reasonable prices.
Instead we got one of the messiest graphics card launch years in recent memory. Confusing pricing, limited stock, shifting launch strategies and plenty of frustration on both the reviewer and consumer side turned what could have been a golden era into yet another round of GPU price drama.
At the center of it all sit two cards that should have been easy recommendations for gamers. AMD’s Radeon RX 9070 XT and Nvidia’s RTX 5070 Ti. Both powerful, both well positioned in their respective lineups and both completely undermined by the way they actually hit the market.
The Radeon RX 9070 XT Launch That Went Off The Rails
AMD’s RDNA 4 based RX 9070 XT was supposed to be a statement card. A return to form with strong performance at a much lower price than Nvidia’s rival offerings. The target price for the 9070 XT was around 599 US dollars which, given its performance, could have made it one of the best value high end GPUs in years.
In reality, everything from pricing to availability was a mess.
First AMD did not release a reference model for reviewers or customers. That meant the only versions of the card available at launch were custom designs from partners like XFX and Asus, each with their own cooler designs and crucially their own price tags.
For reviewers trying to test the card properly this caused chaos. One day a board partner would say their card will be sold at something close to the official MSRP. The next day that same model would be floated at 700 dollars or even 800 dollars with talk of launch pricing that would quickly ramp up once stock hit the shelves.
By the time embargoes were about to lift, pricing information for different RX 9070 XT models was still inconsistent. One minute a card is “reference priced” for review purposes, the next it is classed as a premium model that will actually hit shelves hundreds of dollars higher. Some board partners even reversed decisions after launch, promising MSRP level pricing that never became reality at major retailers.
The end result for gamers was simple. A GPU that looked incredible at 599 dollars routinely showed up in stores for 900 dollars or more during its first few months. The theoretical price to performance crown that AMD could have won was lost the second real world prices drifted so far above the official number.
Nvidia’s Blackwell Cards Were Not Much Better
Nvidia’s early Blackwell generation launches might have looked more organized on the surface thanks to their own Founders Edition models. At least those gave the company a way to demonstrate that its suggested prices were possible in theory.
But the story from the gamer’s side was still rough. Reviewers had to deal with late or unstable drivers and very tight review windows. Early buyers faced black screen issues, cable concerns, low supply and aggressive pricing from board partners and retailers.
Nvidia heavily marketed impressive generation on generation gains, especially with cards like the RTX 5070 which was framed as offering RTX 4090 level performance in some scenarios. The catch was that much of this uplift depended on Multi Frame Generation in specific games instead of raw rendering performance from the silicon itself.
At the same time Nvidia reduced die sizes and complexity in certain tiers compared with the previous generation. That left some enthusiasts feeling like the actual GPU under the cooler was not as ambitious as the marketing suggested and that performance per area of silicon was the main metric Nvidia wanted people to care about.
On the price front, the RTX 5070 Ti officially sat above the RX 9070 XT at 749 dollars. However, street prices told a different story. With demand outstripping supply and little practical control from the GPU vendors over final retail pricing, the 5070 Ti also spent much of its early life at well over 900 dollars in many regions.
Why GPU Prices Got So Bad Again
Both AMD and Nvidia took plenty of heat for this generation and some of it was deserved. Driver issues, confusing performance claims and weird product tiering did not help.
However the worst problems for gamers came from the same familiar forces as previous GPU crises. Supply could not keep up with demand, especially in the early months. Add in the knowledge that many PC gamers are willing to pay big premiums for top tier performance and you get a perfect environment for price gouging.
Board partners saw an opportunity to push their custom models far above MSRP. Retailers, remembering how scalpers cashed in during the pandemic and mining booms, leaned into high pricing while stock was thin. The GPU makers could suggest prices, but they could not realistically force the market to follow them.
The supposed 599 dollar RX 9070 XT and 749 dollar RTX 5070 Ti both hovered around or above 900 dollars for a significant part of the year. Only much later as 2025 went on did availability improve and pricing slowly drift back closer to the original launch figures.
Just as things started to calm down, a new threat to reasonable GPU pricing appeared. The so called RAMpocalypse has sent the cost of memory and storage climbing thanks largely to AI demand. There are even rumors that Nvidia has stopped bundling VRAM with its chips for board partners which could push card production costs higher. Since modern GPUs depend heavily on high speed GDDR memory, any spike in RAM pricing almost inevitably trickles down to gamers.
For now, prices for current gen GPUs have softened from their launch peaks, but the lesson of this generation is clear. Performance gains alone do not define a good graphics card release. Without sane pricing and stable availability, even genuinely excellent GPUs can end up feeling like a disaster for PC gamers.
Original article and image: https://www.pcgamer.com/hardware/graphics-cards/well-this-years-graphics-card-launches-were-a-right-shitshow-eh/
