Nvidia, a leading manufacturer of graphics cards, is not one to be complacent. Even though the RTX 40-series has recently been updated, Nvidia is already focusing on the development of its next-generation GPUs – the RTX 50-series.
Table of Contents
Specifications of RTX 50-series
Pricing and Release Date of RTX 50-series
Architecture of RTX 50-series
Performance of RTX 50-series
Power Consumption of RTX 50-series
While the release of the RTX 50-series GPUs might still be some time away, various speculations and leaks provide us with a glimpse of what to anticipate. Here’s a compilation of all the information we have about Nvidia’s forthcoming generation of graphics cards.
Here are the specifications for the Nvidia RTX 50-series:
- Process Node: TSMC 3nm
- Architecture: Blackwell
- Chip Models: GB202, GB203, GB205, GB206, GB207
- Memory Type: GDDR7
- Maximum Bus Width: 384-bit
- Display Connectors: DisplayPort 2.1, HDMI 2.1
These specifications suggest that the RTX 50-series will be a significant upgrade in terms of performance and efficiency. However, please note that these are rumored specifications and may change upon the official release.
While the release of Nvidia’s RTX 50-series GPUs is still some time away, there are no confirmed specifications for any of the cards yet. However, based on various hardware leaks, we can make some educated guesses. Please note that these are unconfirmed and should be taken with a grain of skepticism until Nvidia officially announces the details.
The successor to the Ada Lovelace architecture is known to be Blackwell, named after the American mathematician David Blackwell. It’s rumored to be manufactured by TSMC using a 3nm process, but it’s not clear whether Nvidia will use one of TSMC’s existing 3nm nodes or a custom one.
The lineup is speculated to include chips ranging from the high-end GB202 (equivalent to RTX 4090) to the entry-level GB207. If this is true, it could mean a significant change, as the AD104 GPU that powers the RTX 4070 may not have a successor in the next generation. This could imply that the RTX 5070 and RTX 5070 Ti might use the GB205 chip.
According to kopite7kimi, a well-known leaker, the new GPUs are expected to support DisplayPort 2.1, which is not available in the Lovelace lineup, and HDMI 2.1. They also suggested that the RTX 50-series cards will feature next-gen GDDR7 memory, and the maximum memory bus width could be 384-bit. This contradicts earlier speculation that the RTX 5090 might have a 512-bit memory bus. Remember, these are all speculations and the actual specifications may vary.
A recent leak from RedGamingTech provides a preliminary look at the potential specifications for the various chips in the RTX 50-series lineup. However, these don’t necessarily correspond directly to the final product specifications. For example, while the GB202 chip is likely to be used in the flagship graphics card, it doesn’t mean that the rumored RTX 5080 will be downgraded to the GB203. In fact, the RTX 4090 didn’t fully utilize its AD102 chip, so we might see a similar scenario with the Blackwell architecture.
According to RedGamingTech, the flagship chip could feature 192 streaming multiprocessors (SMs), an increase from the 144 in the AD102. The GB203, on the other hand, represents a significant reduction, with just 108 SMs (up from 80 in the AD103). This suggests that if the RTX 5090 gets most of those SMs and the RTX 5080 is equipped with the GB203 chip, there could be a substantial performance gap between the two cards, similar to the current generation.
The amount of VRAM in GPUs has been a subject of recent discussion, and RedGamingTech speculates that the RTX 5090 might have up to 36GB of memory. However, these figures are not final, so the actual memory could be 24GB, similar to the RTX 4090.
Lastly, RedGamingTech suggests that the GB207, the least powerful chip in the lineup, will initially only appear in laptops. This aligns with what we’ve seen in the RTX 40-series, where the RTX 4050 has so far only been available in laptops.
At this early stage, the specifics of any individual card are still uncertain, and all of this information is subject to change. Nvidia is likely to release models ranging from the RTX 5060 to the RTX 5090, possibly with some Ti options as well. Hopefully, Nvidia will balance the specifications to offer a diverse range of cards for both enthusiasts and entry-level users.
RTX 50-series: Release Date and Pricing
As of now, there are no specific details from Nvidia regarding the release date of the RTX 50-series. However, most estimates suggest that the launch of the Blackwell architecture, which the RTX 50-series is based on, will occur around the end of 2024 or the beginning of 2025. There are even some tentative whispers about a possible refresh of the RTX 50-series in 2026, but that’s too far into the future to be certain.
Early rumors suggested that Nvidia wasn’t expected to launch the new graphics cards until 2025, which would have given AMD a significant advantage, especially since it’s rumored to launch its RDNA 4 GPUs later this year. However, according to the YouTuber and frequent leaker Moore’s Law Is Dead, Nvidia might not give AMD the breathing room it needs. They mentioned in a recent video that a source at Nvidia told them that “Blackwell is being prepared to be ready to launch in the fourth quarter of 2024,” but only if Nvidia decides to do so. This will depend on whether RDNA 4 will be competitive enough to take away sales from Nvidia during the holiday season at the end of this year, as well as how the sales of the Ada architecture are going around that time.
Regardless, Nvidia is supposedly planning to “make a big deal about RTX 5000 efficiency at CES 2025.” This suggests that the GPUs are likely launching either at the end of 2024 or near the beginning of 2025.
As for the pricing, it’s purely speculative at this point. In the current generation, Nvidia adopted a pricing strategy that can be described as “expensive.” It might continue down this path and push the prices even higher, especially if the demand for AI GPUs remains as high as it is now. After all, the current demand pushed the RTX 4090 way above $2,000, even though it launched at an already high price point of $1,600. This makes the RTX 5090 a concerning prospect. However, Nvidia’s recent price cut in the RTX 40-series Super refresh has given many enthusiasts some hope.
Assuming the flagship 5090 will cost close to $2,000, the rest of the lineup is unfortunately likely to follow with price increases across the board. However, for Nvidia to remain the go-to against AMD, the prices can’t keep rising indefinitely. There is some hope that Nvidia will realize this and keep its pricing more reasonable in the next generation, but it’s too early to tell. Please note that these are all speculations and the actual release date and pricing may vary.
RTX 50-series: Architecture
Nvidia has been quite secretive about the architecture used in the Blackwell chips. However, as the GPUs are expected to be released in about a year, more information will likely become available as the release date approaches. Currently, we only have speculations from various sources, and the information can sometimes be conflicting.
RedGamingTech, in a recent video, discussed the Blackwell architecture extensively. They referred to it as “one of the most influential graphics architectures,” predicting that the RTX 50-series will introduce significant improvements to aspects like path tracing and ray tracing, offering gains for both enthusiast-grade and midrange cards.
According to RedGamingTech, we might see significant architectural changes, including a major redesign of Nvidia’s Streaming Multiprocessors (SMs). They also mentioned the addition of a denoising accelerator, either as a part of the chip or as a function of Nvidia’s Tensor cores. More importantly, RedGamingTech initially suggested that Nvidia might use a multi-chip module (MCM) design. This design approach, where multiple smaller chips are packaged together to form a single, larger, and more powerful processor, could give Nvidia a major advantage, including scalability, higher yields, and more design flexibility.
However, a recent update from the same YouTuber revealed that Nvidia won’t be using an MCM design in Blackwell. Reportedly, Nvidia initially planned to use dual GB202 dies glued together, possibly with some SMs cut, but ultimately decided against it. The YouTuber remarked that issues such as high prices, the latency between the two dies, and various difficulties in getting it to work made Nvidia stick to its previous architecture.
These speculations should be taken with a grain of skepticism. It’s possible that Nvidia may be planning to switch to MCM in the future, but such architectural changes are never made last minute, so that plan for Blackwell may have never existed. However, it’s also possible that Nvidia may introduce architectural changes instead of pushing for top performance to allow the new technology to mature before ramping up the performance in RTX 6000-series graphics cards a few years from now.
RTX 50-series: Performance
As the specifications of RTX 50-series graphics cards are still mostly a mystery, it’s hard to make any accurate predictions as to the performance of these GPUs. However, many have tried, which is why we have some juicy rumors to dig into while we wait for official benchmarks.
According to Moore’s Law Is Dead, the performance uplift between Ada and Blackwell may not be major. The YouTuber’s source mentioned that “Blackwell’s rasterization uplift over Ada will not be as impressive as [from] Ampere to Ada.” However, the source also said that Nvidia could make the RTX 5090 feel like a similar uplift “if it felt threatened.” That seems unlikely, seeing as AMD is reportedly stepping down from making high-end GPUs in the next generation, potentially leaving Nvidia as the only source of high-end graphics cards for the next couple of years.
Based on the above, we might be looking at performance gains along the lines of 30% to 50% for the flagship. Midrange and entry-level cards typically see a smaller boost in performance gen-on-gen, so those might be even less impressive.
However, on the other end of the spectrum is speculation from sources like RedGamingTech. The YouTuber claims in his video that we’re looking at an up to 2x increase in performance between Lovelace and Blackwell. He mentioned that the RTX 50-series should double the ray tracing performance compared to the RTX 40-series, as well as provide a performance boost of up to 2x. RedGamingTech is unsure if this means rasterization, though, so it’s hard to know the metric by which to measure these gains. He does, however, predict clock speeds reaching over 3GHz, which would be a sizable boost over Ada, but also says that this only applies to overclocked models.
In a later video, RedGamingTech added that we might see an up to 60% boost from one flagship to the next, meaning a 60% difference in performance between the RTX 4090 and the RTX 5090. The YouTuber also stressed that Nvidia’s focus was heavily on ray tracing and path tracing, with up to a 2.5x boost in those workloads. Again, approach all of this information with some skepticism.
The only real hint of performance figures we have right now comes from a slide made by Nvidia, but unfortunately, the slide talks about its next-gen high-performance computing (HPC) graphics card used in data centers. The graph, which measures GPU performance in GPT-3 175B inference, shows that the H200 GPU will be up to 18 times faster than the A100 — but that’s not Blackwell architecture yet. B100, the first Blackwell graphics card on the list, offers significantly higher performance, although Nvidia didn’t put a number on it. It looks to be about twice as fast as the H200.
While that’s exciting for those in need of an HPC GPU, gamers and other consumers will need to wait to find out the reality about the capabilities of RTX 50-series GPUs.
RTX 50-series: Power draw
Before the launch of the RTX 40-series, there were numerous rumors about the power consumption of the flagship RTX 4090, with some sources claiming it could reach up to 900 watts. However, these claims turned out to be false, as the RTX 4090 consumes 450 watts, and its connector supports up to 600W, albeit with occasional overheating issues. It’s hard to imagine Nvidia pushing these numbers even higher in the next generation of GPUs.
The RTX 50-series has not been immune to power-related controversies. According to Moore’s Law Is Dead, Nvidia is planning to use a new connector, marking the fourth such change in just three years. They cited anonymous sources, claiming that Nvidia is switching to a 16-pin connector, all dedicated to 12V power delivery. However, many other sources suggest this is unlikely.
Hardware Busters confirmed through their own sources that “no one is aware of a new connector.” Nvidia would need to collaborate with major PSU brands, especially given the issues with the 12VHPWR connector. If these brands are unaware of it, Nvidia might not be implementing these changes in this generation.
Assuming Nvidia continues to use the somewhat controversial 12VHPWR connector, the maximum power consumption will remain at 600W. The flagship RTX 5090 might see an increase in power draw if it offers significantly more performance, but it would still need to leave some room for potential overclocking, so a maximum of 500W seems reasonable.
For the rest of the lineup, Nvidia might aim for more conservative power consumption instead of pushing for higher levels. As NotebookCheck pointed out, Nvidia’s current trend of increasing total board power (TBP) is relatively new, especially for cards like the RTX 4080. Historically, xx80 cards stayed well under 300W, even dipping below 200W at times. In the last couple of generations, the RTX 3080 and the RTX 4080 both pushed the TBP to new heights, each requiring up to 320W.
Given the high power consumption, it doesn’t make much sense for Nvidia to continue pushing for even higher wattages, especially as AMD is likely to be more conservative with its RDNA 4. If Nvidia dials it back a bit, we might see the RTX 5080 with a TBP around 250W to 280W. However, if Nvidia sticks to its current scheme, it might go in the other direction and hit as high as 350W. Please note that these are all speculations and the actual power draw may vary.
Related Posts
OpenAI Unveils SearchGPT: A New Challenger to Google’s Throne
Breaking News: OpenAI, the wizards behind ChatGPT, have just pulled a rabbit out of their digital hat. Meet SearchGPT, the latest
Apple Maps Expands to the Web: Taking on Google with a Bold Move
More than a decade after its debut, Apple Maps is stepping up its game and venturing onto the web, aiming