Do Nvidia’s GeForce RTX 50-series specs dwell as much as the hype?

Nvidia is really masterful at one factor: completely showcasing the graphics playing cards on the launch of a brand new GPU era. This isn’t dangerous in itself; in spite of everything, a fruit vendor on the market additionally advertises his vivid pink apples as an actual deal with.

However apple doesn’t value $2,000 like Nvidia’s new gaming flagship, the GeForce RTX 5090. Let’s take an in depth take a look at the technical specs of the 4 new graphics playing cards — the RTX 5070, RTX 5070 Ti, RTX 5080, and RTX 5090 — so to make the suitable buy resolution in your wants with out having to parse all of the advertising discuss DLSS 4 video games that haven’t even launched but.

After all, you need to anticipate impartial checks to be revealed simply earlier than the 5090 and 5080 go on sale on the finish of January 2025. The launch of the 5070 and 5070 Ti will then observe on the finish of February.

From $2,000: Nvidia GeForce RTX 5090

Have you ever already acquired used to the truth that a single {hardware} element alone now prices as a lot as a high gaming PC used to? Effectively, I haven’t, however that’s how the free market financial system works, and Nvidia has de facto monopolized the graphics card market with an 88% share. So ought to we be grateful that Nvidia affords us by far the quickest graphics card on the planet for a surcharge of “only” $400 over its predecessor, the RTX 4090?

Let’s put it this fashion: if PC gaming is considered one of your favourite hobbies and you’ve got the cash, then go for it.

The RTX 5090 has nearly 33 p.c extra shader models and FP32 computing energy has elevated by nearly 27 p.c in accordance with the information sheet. As well as, there’s over 33 p.c extra VRAM with a wider reminiscence interface (33%) and sooner GDDR7 reminiscence sort (33%). As you possibly can see, the technical information alone guarantees a efficiency improve of round 30 p.c. Added to this are the enhancements to the GPU structure, which also needs to guarantee extra FPS, in addition to the elephant within the room: DLSS 4.

It was predictable that Nvidia would nearly completely present benchmarks with activated DLSS 4 and ray tracing on the launch of the RTX 50-series. Nonetheless, you need to take the expected efficiency improve of one hundred pc and extra with a pinch of salt, as we don’t but know what impact the extra AI frames can have on picture high quality. To not point out that seven gaming benchmarks (chosen by Nvidia) solely mirror a tiny fraction of the gaming market.

The RTX 5090 is the graphics card of alternative in the event you earn your cash with AI improvement and may’t afford an expert card. Nvidia guarantees a efficiency improve of as much as one hundred pc in comparison with the 4090 on this space.

From $999: Nvidia GeForce RTX 5080

In distinction to the 5090, the technical information of the Nvidia GeForce RTX 5080 appears to be like quite modest. In comparison with its predecessor, the 5080 solely has round 10 p.c extra shader models and, in accordance with Nvidia, the FP32 computing energy has elevated by slightly below 20 p.c.

The quantity of reminiscence and connection stay the identical at 16GB and 256 bits. Nonetheless, the RTX 5080 advantages from sooner GDDR7 VRAM (33%) and the MSRP has been decreased by $200.

As a result of small improve in uncooked computing energy, the 5080 isn’t a price-performance champion. This can be completely different for video games that assist DLSS 4. However as I mentioned, solely a small a part of the gaming world helps DLSS 4, so it’s essential to weigh issues up fastidiously. You also needs to be important of Nvidia’s historically “restrained” reminiscence configuration.

Nvidia

Nvidia makes use of numerous methods to maintain reminiscence necessities artificially low, but when the VRAM overflows, this may result in large drops in efficiency. And the reminiscence necessities of future video games will have a tendency to extend considerably, so that you also needs to take a important take a look at the future-proofing of the RTX 5080.

So who should purchase the 5080? An improve from the 4080 is unnecessary, however because the proprietor of a 20 or 30 sequence card, the 5080 is a sexy choice in the event you’re able to fork over $999 for a graphics card.

From $749: Nvidia GeForce RTX 5070 Ti

For my part, the Nvidia GeForce RTX 5070 Ti at the moment affords one of the best price-performance ratio of the brand new 50-series — at the very least on paper. The RTX 5070 Ti already has nearly 17 p.c extra shader models in comparison with its predecessor.

Nvidia

The decisive issue, nevertheless, is the quantity of reminiscence and the reminiscence bus measurement, which has been elevated from 12 to 16 gigabytes and from 192 to 256 bits. There may be additionally GDDR7 VRAM (33% greater than predecessor); and the road worth is a tad cheaper.

This makes the 5070 Ti one of many extraordinarily uncommon Nvidia playing cards with an affordable reminiscence configuration. For my part, this might wind up being the right card from Nvidia for 1440p avid gamers!

From $549: Nvidia GeForce RTX 5070

The worst deal within the new 50-series appears to be like to be the Nvidia GeForce RTX 5070, which affords a modest 4 p.c extra shader models and an solely imperceptibly increased FP32 computing energy of 30.8 TFLOPS in comparison with its predecessor.

Nonetheless, the RTX 5070 solely turns into a pipe-wrecker as a result of insufficient reminiscence gear and connection, with 12 gigabytes and a 192-bit huge interface. Even the sooner GDDR7 VRAM can not make up for this. To place it bluntly: it’s fairly outrageous to equip a graphics card costing over $500 with solely 12 gigabytes.

Nvidia

Additional studying: Better of CES 2025: The PC and residential tech that blew us away