Nvidia is really good at one thing: showing off the graphics cards perfectly when a new generation of GPUs is released. That's not bad in itself; after all, a fruit vendor at a market would also advertise his bright red apples as a real delicacy.

But apples won't be priced at $2,000 like Nvidia's new gaming flagship, the GeForce RTX 5090. Let's take a closer look at the technical specifications of these four new graphics cards (RTX 5070, RTX 5070 Ti, RTX 5080, and RTX 5090) so that you can make the right buying decision for your needs without having to parse all the marketing talk about unreleased DLSS 4 games.

Of course, you should wait until independent test results are released before the 5090 and 5080 become available in late January 2025. The 5070 and 5070 Ti will be available in late February.

From $2,000: Nvidia GeForce RTX 5090

Have you gotten used to the fact that individual hardware components now cost as much as a top-of-the-line gaming PC? Well, I haven't, but that's how the free market economy works, and Nvidia has a de facto monopoly on the graphics card market with an 88% share. So should we be thankful that Nvidia has given us the world's fastest graphics card to date, for a price that is "only" $400 more than its predecessor, the RTX 4090?

Let's put it this way: if PC gaming is one of your favorite hobbies and you have enough money, go for it.

According to the data sheet, the RTX 5090 has almost 33% more shader units and almost 27% more FP32 computing power. In addition, there is more than 33% more VRAM, a wider memory interface (33%), and a faster GDDR7 memory type (33%). As you can see, the technical data alone promises a performance increase of about 30%. Add to that the improvements in the GPU architecture, which should also ensure higher FPS, and the elephant in the room: DLSS 4.

Predictably, Nvidia showed almost exclusively benchmarks with DLSS 4 and ray tracing enabled at the launch of the RTX 50 series. However, you should be skeptical of projected performance increases of 100% or even more, since we don't yet know what effect the additional AI frames will have on image quality. Not to mention that the seven game benchmarks (selected by Nvidia) reflect only a small part of the gaming market.

If you make your money from AI development and can’t afford a professional graphics card, then the RTX 5090 is the graphics card to go for. Nvidia promises up to 100% performance improvement compared to the 4090 in this area.

From $999: Nvidia GeForce RTX 5080

The Nvidia GeForce RTX 5080's specs look pretty average compared to the 5090. The 5080 has only about 10% more shader units than the previous generation, and according to Nvidia, just about 20% more FP32 compute power.

The amount of memory and connectivity remains the same at 16GB and 256-bit. However, the RTX 5080 benefits from faster GDDR7 VRAM (33%) and the MSRP is also $200 lower.

With a modest increase in raw compute power, the 5080 isn't a price/performance champ. The situation might be different for games that support DLSS 4. But as I said, only a small subset of games support DLSS 4, so it's important to weigh the trade-offs carefully. You should also criticize Nvidia's traditionally "restricted" memory configurations.

Nvidia uses various techniques to artificially lower memory requirements, but if VRAM overflows, it can lead to a significant drop in performance. And memory requirements for future games tend to increase significantly, so you should also seriously consider the future development of the RTX 5080.

So who should buy the 5080? There's no point in upgrading from the 4080, but as the owner of a 20 or 30 series graphics card, the 5080 is an attractive option if you're willing to spend $999 on a graphics card.

From $749: Nvidia GeForce RTX 5070 Ti

I think the Nvidia GeForce RTX 5070 Ti is currently the best value for money in the new 50 series - at least on paper. The RTX 5070 Ti has increased the number of shader units by nearly 17% compared to its predecessor.

But the deciding factors are the memory capacity and memory bus size, which has increased from 12 GB to 16 GB and the bus length from 192 bits to 256 bits. There's also GDDR7 VRAM (33% more than the previous generation); and it retails for a slightly lower price.

This makes the 5070 Ti one of the extremely rare Nvidia cards with a reasonable memory configuration. In my opinion, this card may finally become Nvidia's perfect card for 1440p gamers!

From $549: Nvidia GeForce RTX 5070

The worst deal of the new 50 series appears to be the Nvidia GeForce RTX 5070, which has just 4% more shader units and only slightly more FP32 compute power at 30.8 TFLOPS compared to its predecessor.

However, the RTX 5070's lack of memory and connectivity, with only 12GB and a 192-bit wide interface, makes it a pipeline buster. Even faster GDDR7 VRAM can't make up for this. To put it bluntly: a graphics card that costs more than $500 and only comes with 12GB of memory is outrageous.

Source: Content compiled from PCWorld

View more at EASELINK