Nvidia GeForce RTX 5090 could be up to 70% faster than the 4090, but its best chips might be reserved for AI

The Nvidia GeForce RTX 5090 graphics card has been the subject of plenty of rumors since at least last year, and the latest gives us two new surprising details on what we can expect from the possible next-gen graphics card. 

The RTX 5090 most likely will be based on the Nvidia Blackwell architecture and, according to the tech YouTube account Moore’s Law is Dead, could be up to 70% faster than the current-gen RTX 4090 graphics card. This is an absolutely massive leap in performance, which would make child’s play of any of the best PC games currently on the market. 

Previous rumors had put the performance jump for the RTX 5090 at nearly twice as fast as the RTX 4090, so this is one more data point in that bit of speculation's favor. But, as with all rumors, some skepticism is warranted until we can actually measure that performance ourselves. 

This performance boost would likely come from as many as 192 streaming multiprocessors in the RTX 5090 (a 50% increase over the RTX 4090's 128), giving the card 24,576 CUDA cores, 192 ray tracing cores, and 768 tensor cores. In other words, if any of these rumors pan out then this card will be a true behemoth.

However, this spike in speed between the cards would come at a steep price, literally. The same report suggests that the 5090 could end up costing between $2,000-$2,500, with other possible prices including a $1,000 RTX 5080 card, a $700 RTX 5070 card, a $400 RTX 5060 card, and then a budget RTX 5050 Ti card for around $300.

Despite the performance and costs, these might not even be the best cards that Nvidia can offer. There’s another prediction that the card will most likely not have a fully enabled die since Nvidia will almost certainly save its most powerful cards for the booming AI market, the same market that launched the tech giant into the trillion-dollar earning stratosphere last year.

Nvidia is seemingly unchallenged

It’ll be fascinating to see how Nvidia’s focus on AI this generation will affect its graphic card development in this next generation. We might get some obscenely powerful cards coming our way and yet its die won’t even be fully enabled for the sake of the AI market over gamers, which truly boggles the mind.

And there’s the fact that even knowing this, Team Green can charge however much they like for this card if AMD can’t step up to the plate and produce a graphics card of a similar caliber. And with speculation that AMD is not even going to make a play for the premium GPU market with its next-gen RDNA 4 architecture, it doesn't seem likely that anything will be able to restrain the rising prices for Nvidia's graphics cards.

You might also like

• 16-pin connectors could be mandatory for all Nvidia RTX 5000 chips
• Nvidia 5000 series GPUs might use multi-chiplet design
• Nvidia's RTX 5000 GPUs are a confusing proposition for gamers