As we get deeper into 2024, the news and rumors about Nvidia Blackwell are starting to pick up as we learn more about Nvidia Lovelace’s architectural successor.
This includes both data center and industrial scale chips, like the ones fueling the current AI boom, as well as the next-generation Nvidia RTX 5000 series. And, unlike previous generations of architecture, Nvidia is much more open about it as it continues to promote its AI hardware. However, the company has been fairly tight-lipped about its plans for its GeForce line, including its flagship Nvidia RTX 5090 and enthusiast Nvidia RTX 5080 graphics cards.
However, the rumor mill continues to churn, and as we get closer to the expected launch late this year, we’re hearing more about what to expect from the next generation of Nvidia’s top graphics cards, and we’ve even gotten some insight into the architectural changes that will underlie it. Here’s everything we know so far.
Nvidia Blackwell: cut to the chase
- What is it? Nvidia’s rumors are about its next-generation GPU architecture
- How much is it? Unknown at this time, but will scale from low budget graphics cards to AI hardware for data centers.
- When can I get it? The earliest we’d expect to see Nvidia Blackwell would be mid-to-late 2024, with Blackwell’s data center products likely rolling out before the consumer line.
Nvidia Blackwell: Latest news
Read more of the latest Nvidia Blackwell news…
Nvidia Blackwell: release date
All we can say about Nvidia Blackwell right now is that it looks to release in mid-to-late 2024 for its data centers and industrial Blackwell GPUs, and early 2025 for the bulk of its consumer products, at least according to Nvidia’s own roadmap. However, there is a lot of talk that the Nvidia RTX 5090 and possibly even the RTX 5080 will arrive before the year is out.
Nvidia Blackwell: specs
We know next to nothing about what kind of architecture Nvidia Blackwell will present, including which process node it will use, who will fabricate the chips using the architecture, or what innovations Nvidia might roll out with it.
We expect more mature ray tracing and tensor cores, and it’s almost guaranteed that Nvidia will focus on the latter rather than the former, since it’s the tensor cores that are powering Nvidia’s current dominance of AI workloads.
We’ve gotten hints that Nvidia Blackwell will use, at least in part, a multi-chiplet module design, which would be a first for Nvidia. The rumor mill doesn’t specify whether both commercial and consumer Blackwell GPUs will use MCMs, but it’s an exciting prospect nonetheless and means Nvidia is catching up with AMD and Intel, who have already adopted the design paradigm in their own processors.
There have been some rumors about the specific GPUs that will actually run the Nvidia Blackwell architecture. The designations for consumer GPU variants appear as follows:
- GB202: probably in the Nvidia RTX 5090 and Nvidia RTX 5090 Ti
- GB203: probably in the Nvidia RTX 5080 and Nvidia RTX 5080 Ti, but possibly also in the Nvidia RTX 5070 Ti.
- GB205: Probably in the Nvidia RTX 5070 and RTX 5060 Ti, and possibly in the Nvidia RTX 5070 Ti
- GB206: Probably 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.
We’ve also seen some alleged specs for an RTX 5090 from a fairly reliable leaker on the Chinese-language Chiphell forum. According to a now-deleted report, the RTX 5090 will feature some impressive spec upgrades over the RTX 4090:
| Spec | RTX4090 | RTX5090 |
|---|---|---|
| Streaming multiprocessors | 128 | 192 |
| CUDA cores | 16,384 | 24,576 |
| Ray Tracing Cores | 128 | 192 |
| Tensor nuclei | 512 | 768 |
| Boost clock | 2.52GHz | 2.9GHz |
| L2 cache | 72MB | 128MB |
| Memory bandwidth | 1,008 GB/sec | 1,532 GB/sec |
If these specs come true, it should give the RTX 5090 an absolutely massive gen-on-gen boost, with the same post detailing the specs claiming the RTX 5090’s performance was 1.7 times faster than the RTX 4090, which is downright mind-boggling.
Nvidia Blackwell: what to expect
Being so far away from a release, it’s obviously hard to say what we can expect from the new architecture. The only thing that’s certain is that there will be more developments, leaks, and rumors as we get closer to release. So keep an eye on Ny Breaking for the latest news on Nvidia’s next-generation graphics architecture.