The Nvidia 5000-series graphics cards have been the talk of the tech world for some time now, and one of the most widely speculated-about cards has been the RTX 5090, the flagship of the lineup and likely the best graphics card on the market once it eventually launches.
With how well the current-gen flagship Nvidia RTX 4090 was reviewed (we cited it as one of the best graphics cards ever released), there are high expectations for next-gen cards, and the RTX 5090 in particular bears the brunt of that. The Nvidia Lovelace generation as a whole has been a mixed bag in terms of performance and pricing, so that adds even more pressure to the 5090 to succeed in both aspects.
However, right now it’s still too early to say whether it can live up to these expectations, with plenty more to learn about until it’s officially revealed by Nvidia. For now, let’s go over what we do know thanks to all the rumors and leaks.
Nvidia RTX 5090: Cut to the chase
- What is it? Nvidia’s rumored flagship for its next generation of RTX graphics cards
- How much does it cost? Unknown at this time, but will likely scale similarly to Nvidia Lovelace GPUs in price
- When can I get it? The earliest we expect to see the Nvidia RTX 5090 would be late 2024 or early 2025
Nvidia RTX 5090: Latest news
Nvidia RTX 5090: Release date
Though there’s no official release date for the Nvidia RTX 5090, thanks to the tech rumor mill we do have an idea as to when the card could come out. And since the global chip shortage has eased up, there aren’t any predicted delays for the next generation chip either.
All signs seem to point to late 2024 or early 2025 at the absolute earliest, which ties into the normal 18 to 24-month release cycle that Nvidia normally abides by.
(Image credit: Nvidia)
Nvidia RTX 5090: Specs
There has been some information coming out concerning the Nvidia RTX 5090’s specs, with the most important being that it’s likely based on Nvidia Blackwell architecture, the rumored early code name for the Nvidia Lovelace successor. It’s also been called Lovelace-Next by Nvidia in official presentation materials.
We also learned that the 16-pin 12VHPWR connector could be a standard for the 5090 and other 5000-series cards, meaning even third-party board partners would need to use the it. It’s also rumored to feature a 512-bit memory bus that is 33% wider than the one on Nvidia’s RTX 4090, the memory bandwidth will feature a 77% boost compared to the 4090, and 28Gbps GDDR7 memory modules which would be 33% faster than the 21Gbps memory modules in the RTX 4090.
{ window.reliablePageLoad.then(() => { var componentContainer = document.querySelector(“#slice-container-newsletterForm-articleInbodyContent-z5RYWf4MQ53riKdn6WLYtH”); if (componentContainer) { var data = {“layout”:”inbodyContent”,”header”:”Get daily insight, inspiration and deals in your inbox”,”tagline”:”Get the hottest deals available in your inbox plus news, reviews, opinion, analysis and more from the TechRadar team.”,”formFooterText”:”By submitting your information you agree to the Terms & Conditions and Privacy Policy and are aged 16 or over.”,”successMessage”:{“body”:”Thank you for signing up. You will receive a confirmation email shortly.”},”failureMessage”:”There was a problem. Please refresh the page and try again.”,”method”:”POST”,”inputs”:[{“type”:”hidden”,”name”:”NAME”},{“type”:”email”,”name”:”MAIL”,”placeholder”:”Your Email Address”,”required”:true},{“type”:”hidden”,”name”:”NEWSLETTER_CODE”,”value”:”XTR-D”},{“type”:”hidden”,”name”:”LANG”,”value”:”EN”},{“type”:”hidden”,”name”:”SOURCE”,”value”:”60″},{“type”:”hidden”,”name”:”COUNTRY”},{“type”:”checkbox”,”name”:”CONTACT_OTHER_BRANDS”,”label”:{“text”:”Contact me with news and offers from other Future brands”}},{“type”:”checkbox”,”name”:”CONTACT_PARTNERS”,”label”:{“text”:”Receive email from us on behalf of our trusted partners or sponsors”}},{“type”:”submit”,”value”:”Sign me up”,”required”:true}],”endpoint”:”https://newsletter-subscribe.futureplc.com/v2/submission/submit”,”analytics”:[{“analyticsType”:”widgetViewed”}],”ariaLabels”:{}}; var triggerHydrate = function() { window.sliceComponents.newsletterForm.hydrate(data, componentContainer); } if (window.lazyObserveElement) { window.lazyObserveElement(componentContainer, triggerHydrate); } else { triggerHydrate(); } } }).catch(err => console.log(‘Hydration Script has failed for newsletterForm-articleInbodyContent-z5RYWf4MQ53riKdn6WLYtH Slice’, err)); }).catch(err => console.log(‘Externals script failed to load’, err)); ]]>
Get the hottest deals available in your inbox plus news, reviews, opinion, analysis and more from the TechRadar team.
We recently heard rumors that Nvidia would finally be switching some of its Blackwell GPUs over to a multi-chiplet module (MCM) design, following in the footsteps of AMD and Intel. Whether this will include the Nvidia 5000 series GPUs isn’t clear, however, since the rumors only specified the GB100 GPU, which is a commercial-grade chip for servers, data centers, and industrial use.
Still, an MCM Nvidia GPU could provide a big boost to performance if done properly, and given that archrival AMD is already using MCMs in its GPUs, Nvidia can’t afford to get left behind here.
We’ve also seen some purported specs for an RTX 5090 from Chiphell forum user Panzerlied, a fairly reliable hardware leaker. According to a now-deleted post, the RTX 5090 will boast some impressive spec upgrades over the RTX 4090:
Swipe to scroll horizontally
| Spec | RTX 4090 | RTX 5090 |
|---|---|---|
| Streaming Multiprocessors | 128 | 192 |
| CUDA Cores | 16,384 | 24,576 |
| Ray Tracing Cores | 128 | 192 |
| Tensor Cores | 512 | 768 |
| Boost Clock | 2.52 GHz | 2.9 GHz |
| L2 Cache | 72MB | 128MB |
| Memory Bandwidth | 1,008 GB/s | 1,532 GB/s |
If these specs pan out, this should give the RTX 5090 a massive gen-on-gen boost, with the same post that detailed the specs claiming that the RTX 5090’s performance was 1.7 times faster than the RTX 4090, which is downright wild.
Beyond this, there isn’t much else that we know about these GPUs, such as what process node they will be fabricated on, who will be doing the fab (though it’ll almost certainly be TSMC, as was the case with Nvidia Lovelace), and what kind of core counts and clock speeds we should expect.
Nvidia RTX 5090: What to expect
What we can expect first and foremost is that, if the rumors concerning the 5090 are true, there will be plenty more information to come before its eventual release. Of course, there could be more released about the current-gen Lovelace cards as well, like a budget-minded RTX 4050.
As we learn more we can start to build proper anticipation around the release, though it is still a long way off all of this should be taken with a hefty grain of salt for the time being.