It has been over a month since NVIDIA unveiled its 50 series RTX GPUs at CES 2025, and it’s clear that these graphics cards have failed to impress. The problems plaguing these GPUs include manufacturing issues, misleading marketing around the 5070, and minimal performance improvements compared to the 40 series. This launch might be one of the worst in recent memory for NVIDIA. The most disheartening aspect, however, is the repetition of past mistakes.

Reflecting on the past few weeks, I am reminded of an article I wrote in 2022. In it, I discussed my experience buying an RTX 3070 in August of that year, just over two years before the launch of the GeForce 40 series. Many readers pointed out that my predictions about the crypto market’s impact on future pricing were incorrect, but I believe my underlying reasoning was sound. I purchased the 3070 when I did because I anticipated that NVIDIA’s new GPUs would be more expensive than their 30 series counterparts, and finding a new card at launch for a reasonable price would be nearly impossible. As it turned out, I was correct: the 40 series cards were difficult to find at launch.

Fast-forward to the present, at the beginning of a new GPU generation, and we find ourselves in a similar situation. For regular consumers looking to upgrade their gaming PCs, it’s not only impossible to find a 50 series GPU in stock but also nearly every model is significantly overpriced compared to NVIDIA’s suggested price. This time, there’s no pandemic to blame, so what’s causing these issues?

The root of the problem lies in the precedent set by trends during the COVID-19 pandemic. Since 2020, some consumers have been willing to pay exorbitant prices for new NVIDIA cards at launch. Rather than allowing scalpers to profit from this behavior, NVIDIA’s partners have decided to capitalize on it themselves. As a result, the company’s GPUs are now priced as high as the market will bear, and they will continue to carry these inflated prices as long as supply remains limited.

It’s time to acknowledge that the business model underlying the GPU industry is broken. For those unfamiliar, AIBs (add-in board partners) like ASUS, Gigabyte, XFX, and Zotac produce most of the GPUs available from NVIDIA and AMD. In the past, this model led to differentiation between cards in the same tier. However, those days are long gone.

Modern AIB GPUs often feature minor overclocks or an additional fan, but they are essentially the same product. Whether you purchase a 5070 directly from NVIDIA or one of its partners, it will still be a 5070. The only noticeable difference might be a few percentage points in performance if you opt for a model with an overclock or enhanced cooling. Despite this, pricing can vary significantly between different AIBs and even within a single company’s lineup.

Digital Foundry’s Alex Battaglia aptly described the issue in one of the outlet’s recent DF Direct episodes. “The AIB model seems, for many cards, like a middleman that exists solely to extract more value from consumers without providing valid differentiation for the price increase.”

Consider the 5070 Ti, for instance. According to NVIDIA’s website, the starting price is $749. However, a visit to retailer websites reveals a different story. On Newegg, for example, there isn’t a single 5070 Ti listed for $749. The most affordable option starts at $799, with every other model priced above $830, and one even listed at $920. The value proposition of the 5070 Ti is significantly worse at $900 compared to $750.

Newegg briefly blamed tariffs for the inflated prices of NVIDIA’s new 50 series GPUs, but this explanation would only be plausible if the price increase was around 10 percent.

NVIDIA seems uninterested in addressing the issue in a meaningful way. The company’s best solution so far has been a program that allows an unspecified number of US customers to purchase a 5090 or 5080 directly from the company. Provided AMD’s new $549 Radeon 9070 and 9070 XT cards deliver the promised performance, they may pressure NVIDIA to price the 5070 and 5070 Ti more competitively. However, it remains to be seen if AMD can maintain stock and keep the cards at their suggested retail price.

Pricing for this generation’s GPUs will eventually normalize, but unless there’s a fundamental change in the industry’s business model, we can expect this cycle to repeat. Just because the industry has always operated this way doesn’t mean we should accept the status quo, especially as it worsens with each launch cycle.

If you purchase something through a link in this article, we may earn a commission.