The GeForce RTX 4070 officially turned three this April, and yet it feels as if it were yesterday when the Ada Lovelace GPU was announced. This isn't nostalgia talking, not in the slightest. Rather, it feels like there has been no significant shift in the experience of owning a GPU, no missing features, and alongside it, no reason to eye a potential upgrade over the past three years.

DLSS 4.5 brought the second-generation transformer model that addressed most shortcomings of Super Resolution, whereas Multi Frame Generation took care of the frame output ceiling. All of this being available on a card released three years ago could mean Nvidia has built a true future-proof lineup with the RTX 40 series, but it could also mean that consumer hardware is no longer seeing innovation at the speed of light. Either way, an upgrade isn't on the cards, no matter which way I approach the argument. That, however, isn't great news. Here's why.

Ada Lovelace might be the Pascal series of the 2020s

Turing and Ampere did not age the same way

Looking back, it's almost amusing how many of the features that form an integral part of our gaming experience arrived in an unrefined form. When real-time ray-tracing debuted with the Turing lineup in 2018, Cyberpunk 2077 brought the GPUs that released with it to their knees, to the extent that most people without an RTX 2080 Ti would not even enable it during gameplay. Ampere GPUs slightly improved the experience, but it wasn't until the trifecta of Ada Lovelace, DLSS 3, and the third-generation ray-tracing cores finally made it usable.

On the subject of Super Resolution, DLSS followed a similar course. Each generation subsequently refined the trajectory until it became difficult to imagine modern AAA gaming without it. Every feature that Turing and Ampere introduced in a rough, proof-of-concept form was solidified with Ada, and it's hard to argue otherwise. The 40-series represented the amalgamation of the flagship features of GPUs released five years prior, with a sufficiently advanced architecture to actually deliver the said features without them compromising the key aspects of user experience.

DLSS 4.5 was the final piece, and the 40-series was ready for it

The most comprehensive GPU software stack, on hardware built for it

When DLSS 4.5 came out this year, it initially felt like a win for every RTX-series GPU owner. As Nvidia's headline item at CES 2026, who could help be excited about the second-gen transformer? The promises were huge enough and worth taking a second look. Improved temporal stability and expanded access to Multi Frame Generation sounded exactly like what the successor to DLSS 4.0 needed. It felt as though Nvidia opened and looked through the fifty-thousand Reddit threads complaining about DLSS, ghosting, smearing, foliage shimmering, artifacting, and input lag, and delivered on every single gripe enthusiasts had that made them avoid DLSS like the plague.

Then, the promises met the expectations halfway. While older RTX cards could "technically" support the new feature suite, only two families of GPUs possessed the native FP8 capabilities that allowed the transformer models to run without a ginormous overhead that all but shattered immersion in gaming. The first, of course, was the Blackwell series. The second was the Ada Lovelace lineup. Despite being released 3 years before the second-generation transformer, the 40-series didn't need an entire architectural revamp to deliver the temporal stability and image quality enhancements that made DLSS 4.5 worthwhile. It feels now that, from the day it was built, it was built to last.

For the first time in the history of consumer graphics...

The future seems less exciting than the present

The RTX 40-series sits at a unique intersection that is difficult to manufacture by design. By that, I mean that in this generation, raw rasterization performance holds up against current titles, a full DLSS stack runs natively on hardware it was built for, and access to MFG exists without the ecosystem dependency that makes the Blackwell series feel premature. It is the last GPU generation where the hardware itself was the headline, rather than the software features built around its limitations.

For all those reasons, it seems as though Ada Lovelace is the perfect sweet spot between raw rasterization capabilities and AI-driven rendering advancements. The same cannot be said about Blackwell and the upcoming era of neural rendering.

In this paradigm, Blackwell seems to arrive at the beginning of a structural shift in rendering pipelines that software itself has not caught up to. Perhaps the RTX 60-series will be built to leverage those advancements fully, but for now, the GPU family that delivers the most value per dollar and just the right mix of usable features and rasterization capabilities, beyond question, is Ada Lovelace.

The 50-series has been a tough sell, and the 40-series is the reason why

For as long as I've been building PCs, new GPU generations always felt exciting because they promised something my current hardware simply couldn't do. With Blackwell, the feeling seems to have disappeared. The RTX 40-series sits in a spot where the hardware remains powerful, the software stack has continued to improve three years on, and the compromises are few enough that upgrading feels like a practice in chasing diminishing returns. As an enthusiast, that's reassuring. As an analyst, that's unfortunate to witness.