These 6 features make Nvidia’s RTX 40 series better than the 50 series

Unless you’ve been living under a rock, Nvidia’s RTX 50 series launch has been a disappointment. The gen-on-gen gains are abysmal, the paper launch means you can’t find anything in stock, and the street prices are off the charts. If you don’t count Nvidia’s Multi Frame Generation as true performance, which many don’t, then there seem to be few reasons to buy an RTX 50 series graphics card.

In contrast, the RTX 40 series looks better than ever. The previous-gen cards seem to be aging well, thanks to similar performance to the latest GPUs, access to many DLSS 4 features, lower TDPs, better prices, and the absence of some less-than-desirable issues seen on the Blackwell cards.

6

32-bit PhysX support

For fans of the classics

In a silent move, Nvidia removed support for 32-bit PhysX applications from the RTX 50 series GPUs. What this means is that some older games that used the 32-bit version of PhysX will now have to rely on the CPU instead to power PhysX effects such as dynamic debris, fluid simulations, and cloth physics. Needless to say, the CPU-powered results are sub-par, forcing some users to resort to a secondary 30 series GPU to handle the PhysX calculations.

Nvidia GPUs belonging to the RTX 40 series and older still support 32-bit PhysX. Unfortunately, the latest and greatest Nvidia GPUs are lagging behind the previous-gen cards in an area that might be important for many gamers. After all, the list of affected games includes beloved titles like the Batman: Arkham series, Borderlands 2, Assassin’s Creed IV: Black Flag, Mafia II, and Metro 2033.

Dropping support for older applications and APIs is natural, as time goes by, but many gamers aren’t aware of this downside of the Blackwell cards. This might be another reason to hold on to your older graphics card, especially if you still enjoy a lot of the older PhysX-powered games on your PC.

Related

7 older GPUs to buy instead of the RTX 50 series

The latest Nvidia GPUs might be underwhelming, so here are 7 current and previous-gen GPUs you should consider instead

5

Lower power consumption

For virtually the same performance

Another reason you might want to consider an RTX 40 series graphics card is the significantly lower TDP on many of these models. For instance, the RTX 4090 sports a 450W TDP, whereas its successor, the RTX 5090, features a massive 575W TDP. Similarly, the RTX 5080 has a TDP of 360W, which is markedly higher than the 320W TDP of the RTX 4080 and RTX 4080 Super. The difference in TDP going down the product stack isn’t huge, but you’ll still find it lower on the RTX 40 series.

The fact that you’re barely getting 10-15% more raw performance with the new RTX 5080 and RTX 5070 Ti compared to their previous-gen counterparts adds to the TDP argument. If the older cards can match the new ones in performance (in terms of actual FPS, they can) at lower TDPs, then why would you want a hotter card warming up your desk and room?

Even the $2,000 RTX 5090 performs less than 30% faster than the RTX 4090 on average, and it consumes up to 40% more power than the latter to do so. It seems the architectural improvements with Blackwell aren’t anywhere close to what we used to see before, and the company is hedging on AI-generated frames to become the future. If you’re not a fan of that, you might be more interested in grabbing an RTX 40 series GPU instead.

Related

Nvidia GeForce RTX 5080 vs 4090: The old flagship still rules (but you can’t get it any more)

The incoming RTX 5080 doesn’t quite manage to beat the outgoing flagship

4

Lower prices

If you can find one in stock

This one’s trickier since the ongoing supply issues seem to be affecting not only the RTX 50 but RTX 40 series cards as well. Finding an RTX 5090 or RTX 5080 in stock at reasonable prices is near-impossible, but each of the RTX 40 series cards is also suffering from the same fate. Resellers and scalpers are busy making bank by charging twice or thrice the price of the Blackwell cards, but the alternative of buying an RTX 40 series card isn’t that simple either.

That said, when you finally find these GPUs in stock, you’re more likely to find, say, an RTX 4070 Ti Super at or around the MSRP than the RTX 5070 Ti. As I mentioned before, the RTX 50 series cards don’t perform much faster than the RTX 40 series models, so the lower prices on the latter are a huge draw for gamers who are out to buy a new GPU.

The supply constraints might take months rather than weeks to get fixed, so there’s probably some waiting in your fortune at the moment. However, you have greater chances of finding RTX 40 series cards in stock at offline stores than a new Blackwell GPU. If you have a Micro Center near you, you might be able to find an RTX 4070 Super in stock.

Related

Buying an RTX 40 series GPU? These are the only 6 GPUs you should upgrade from

If you own one of these older GPUs, an RTX 40 series card is a worthwhile upgrade

3

Access to many DLSS 4 features

You’re not left out in the cold

Nvidia’s DLSS 4 features were a huge part of its RTX 50 reveal, and many thought that it would limit most of the advancements to the new Blackwell cards. However, many of the DLSS 4 features, such as the new Transformer model used for DLSS Super Resolution, are available on the RTX 40 series GPUs too (and the 30 and 20 series also). The RTX 40 series already supports Frame Generation (2x instead of 4x), DLAA, and Ray Reconstruction.

Nvidia also announced that the RTX 40 series will soon receive a new feature called Smooth Motion, similar to AMD’s Fluid Motion Frames. This feature is already available on the RTX 50 series cards, and the RTX 40 release is being worked on. You’d also be happy to know that Nvidia might even bring Multi Frame Generation to the RTX 40 series, according to comments made by Bryan Catanzaro. Whether this happens or not is unclear, but it’s surprising Nvidia isn’t gatekeeping MFG to Blackwell.

Hence, the RTX 40 series isn’t too far away from the RTX 50 series even in the software department. If you have a 1440p, 144Hz-180Hz monitor, like most gamers, the 2x frame generation available on the RTX 40 series GPUs is enough to enjoy a high-refresh-rate experience. Plus, you’ll be able to keep the artifacts and latency penalties to a minimum by avoiding the 3x and 4x frame generation seen on the RTX 50 series GPUs.

Related

3 features RTX 40 owners are getting in the wake of the RTX 50

It’s not all about the new generation.

2

No defective silicon

Missing ROPs and bricked GPUs? Thankfully not

In a bizarre development, many RTX 5090, RTX 5080, and RTX 5070 Ti GPUs have been found to be missing Raster Operating Units (ROPs). Nvidia acknowledged that less than half percent of RTX 5090 and RTX 5070 Ti GPUs have been found using misconfigured GB202 and GB203 GPU dies, which had fewer ROPs than marketed. Nvidia is obviously offering replacements to the affected consumers, but this issue adds to Nvidia’s already huge list of problems.

Reports of Blackwell GPUs bricking themselves, not being recognized on consumer systems, and getting black screens have been circulating for a while now. Nvidia is still investigating the reason behind the problem, and a vBIOS update in the future to fix the issue seems likely. Users have reported using workarounds like locking the refresh rate to 60Hz or downgrading to PCIe 4.0 from PCIe 5.0 to overcome the issue, but these are just temporary measures. Nvidia needs to release a permanent solution as soon as possible.

Related

6 reasons to pick Intel or AMD for your next GPU instead of Nvidia’s RTX 50 series

Nvidia’s RTX 50 series might be hot right now, but there are enough reasons to consider an AMD and Intel GPU

1

No melting connectors

The bare minimum

In an unfortunate repeat of the 2022 fiasco, a handful of users have reported that the power connectors on their RTX 5090 and RTX 5080 graphics cards are melting. The revised 12V-2×6 connector, which was used to overcome the issues with the earlier 12VHPWR connector, is the one at the center of the scandal this time around. Although the new connector managed to fix the problem a few years ago, it seems it can’t handle the RTX 50 series GPUs.

Users buying an RTX 40 series graphics card will not have to go through this ordeal, since there haven’t been any recent reports of this happening on the previous-gen cards. Considering Nvidia still hasn’t worked out a fix for the melting connectors on the RTX 50 series cards, it might be a while till we see a clarification from Nvidia or a new connector.

Related

Why doesn’t Nvidia stick to 8-pin PCIe cables?

Nvidia’s 12VHPWR woes still exist, making everyone think why not use 8-pin cables like before?

The latest isn’t always the best

Nvidia might be trying to sell you the new RTX 50 series graphics cards, but the previous-gen RTX 40 models aren’t too different. Considering the lower prices & power consumption, virtually the same raw performance, access to almost all the DLSS 4 features, and the lack of melting connectors and black screens, the RTX 40 series GPUs seem better in almost every department. The only challenge is to find one in stock at a reasonable price.