Nvidia 5000 series: Rumors, possible GPU specs, and everything we know
Rumors about the upcoming Nvidia 5000-series GPUs have been circulating for a while now. As 2024 rolls around, we’re hearing more and more about them as the rumor mill starts churning faster and faster.
Current Nvidia Lovelace RTX graphics cards have mixed performance. The enthusiast and premium mid-range models have generally performed well, but their high prices have left many gamers unable to afford the more expensive cards.
It’s no surprise, then, that there’s a lot of excitement, but also caution, about what to expect from Nvidia’s top-of-the-line Blackwell-generation graphics cards, the next-generation GPUs (led by the RTX 5090) that may — or may not — hit the market later this year.
In this article, we’ll be collecting all the major release date rumors and other news circulating in the run-up to the Nvidia 5000 series announcement. We’ll keep you updated with any new details as we hear them.
Nvidia 5000 Series: Get straight to the point
- What is it? Nvidia’s Next-Generation RTX Graphics Card Rumors
- How much is it? Currently unknown, but will likely be similar in price to Nvidia Lovelace GPUs
- When can I get it? The earliest we expect the Nvidia RTX 5000 series is late 2024 or early 2025
Nvidia 5000 Series: Latest News
Load more of the latest news about the Nvidia RTX 5000 series…
Nvidia 5000 Series: Release Date
Obviously, we don’t know for sure when Nvidia will release its next generation GPUs, but we can make some educated speculation at this point based on past experience and rumors that have appeared in the public debate.
First, Nvidia typically has an 18-24 month release schedule for its graphics cards. As the global chip shortage eases, we don’t expect any delays that deviate from this schedule.
That would put the Nvidia 5000 series release sometime in late 2024 or early 2025 at the latest. Historically, the rumor mill has been leaning toward a late 2024 release, but more recent speculation is starting to point to an early 2025 launch (likely at CES). Either way, rumors are also pointing to a pair of early Blackwell graphics cards – the RTX 5090 and RTX 5080 (though the latter may arrive just before the flagship).
Nvidia RTX 5000 Series: Specifications
The Nvidia 5000 series RTX GPUs are based on Nvidia’s Blackwell architecture.
According to a leak on Chinese hardware forum Chiphell, which was backed by noted online leaker Kopite7kimi, Nvidia’s next-gen GPU lineup will be undergoing something of a numbering shakeup. The supposed GPU variants, which will form the core of both Nvidia graphics cards and the best gaming laptops featuring next-gen Nvidia GPUs, will include the following:
- GB202: likely in the Nvidia RTX 5090 and Nvidia RTX 5090 Ti
- GB203: likely in the Nvidia RTX 5080 and Nvidia RTX 5080 Ti, but possibly also in the Nvidia RTX 5070 Ti.
- GB205: Likely in the Nvidia RTX 5070 and RTX 5060 Ti, and possibly in the Nvidia RTX 5070 Ti
- GB206: Likely in the Nvidia RTX 5060 and possibly the Nvidia RTX 5060 Ti
- GB207: Probably reserved for the Nvidia RTX 5050 and Nvidia RTX 5050 Ti
In the past, we’ve heard rumors that Nvidia would finally switch some of its Blackwell GPUs to a multi-chiplet module (MCM) design, following in the footsteps of AMD and Intel. Whether this will also include the Nvidia 5000-series GPUs, however, isn’t clear, as the rumors have only specified the GB100 GPU, a commercial chip aimed at servers, data centers, and industrial uses.
At this point, it seems unlikely that this will happen for GeForce cards, but still, an MCM Nvidia GPU could deliver a significant performance boost if done right. And with arch-rival AMD already using MCMs in its GPUs, Nvidia can’t afford to be left behind.
We’ve also seen some alleged 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:
| Specification | 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 | 2.52GHz | 2.9 GHz |
| L2 cache | 72MB | 128MB |
| Memory bandwidth | 1.008 GB/s | 1.532 GB/s |
If these specs are accurate, this should give the RTX 5090 a huge boost over the previous generation. The same report detailing the specs also claimed that the RTX 5090 had 1.7x faster performance than the RTX 4090, which is nothing short of staggering.
Other rumors have pointed to a slightly smaller upgrade for the RTX 5090, around 50% or 60%, but let’s be honest: that would still be a huge upgrade.
We’re also expecting a 28GB of VRAM loadout (with a 448-bit memory bus), in case you had any doubts about just how big Blackwell’s flagship GPU might be.
With a potential release drawing closer, we expect to hear more spec rumors soon, and we’ll update you on the latest updates as they become available.
