•Today marks the fourth anniversary of Nvidia’s GeForce RTX 4090, the enthusiast-level graphics card from the previous generation that launched on October 12 with a launch MSRP of $1,499.
The landscape for high-end GPUs has shifted dramatically since then. The cheapest RTX 5080 currently available now costs more than the 4090’s original launch price, and its direct successor, the RTX 5090, has become so scarce that PC Gamer stopped tracking it as third-party listings are priced at levels the publication is not comfortable recommending.
Rumours suggest the GB202 silicon behind the 5090 may instead be diverted to data center and professional GPUs. According to a report for PC Gamer, this has prompted reflection on what the RTX 4090 represented in an era when enthusiast hardware felt more attainable.
The publication recently retrieved its review sample of the card from its kit cupboard—a GPU described as looking “outrageous” even today. Its specification sheet remains substantial: 24 GB of GDDR6X memory, 76.3 billion transistors, and 16,384 CUDA cores, a 56% increase over the prior-generation card it replaced.
The full spec lists 128 shader modules, 128 ray tracing cores, and 512 Tensor cores. By modern standards it remains “a giant of a graphics card,” and the report notes that many gamers still run one after paying what were then considered exorbitant prices.
In hindsight, buying at launch was reportedly a smart move for those who could afford it. PC Gamer’s Dave flagged the prohibitive pricing in his review but praised its gaming performance, which also offered readers their first proper look at DLSS Frame Generation in its early 2x form.
The RTX 5090 has been a more controversial GPU from the outset. It launched with a $1,999 MSRP and rose well above that in retailer listings almost immediately, making it harder to justify on price-to-performance grounds despite outperforming the RTX 4090 considerably.
The card’s Multi Frame Generation support has also divided gamers’ opinions. Announced alongside the RTX 5080 at CES 2025 in Las Vegas, the feature boosts frame numbers substantially but reignited a “fake frames” debate surrounding older frame-generation implementations.
While the RTX 5090 remains a powerful GPU in its own right, its pricing reportedly pushed many gamers away—particularly those who already owned an RTX 4090 and saw little reason to upgrade from what was still a “monstrous” card.
The report suggests this may be the 4090’s legacy: possibly the last halo GPU that made sense for a typical consumer. It touched on the concept of futureproofing, which the author uses with caution, noting the card remains capable for the foreseeable future while later cards drift further out of reach.
Whatever an RTX 6090-equivalent may bring, the report speculates it will likely be priced ludicrously and packed with Tensor cores to support more Multi Frame Generation features. Nvidia’s focus appears to have shifted toward supplementary hardware over raw performance uplift.
The four-year anniversary of Nvidia’s now-legendary 40-series flagship prompts nostalgia for a time when GPUs were simpler, higher-performance machines available at prices somewhat within reach—unusual given the $1,499 price tag and the state of graphics card pricing in 2026.
1 comment