RTX 50 series release window speculation – Nvidia 5000 series latest news, specs, and rumors

Everything we know about the upcoming RTX 50 series GPUs

Table of Contents Table of Contents

Looking for the latest on the RTX 50 series then you can find all you need to know right here. With Nvidia’s series 5000 series expected release date information, the latest specs, and rumors on the generation of cards. With plenty of expectations behind them and some disappointment from the 40 series, we give you everything we know about the Blackwell range of GPUs.

The highly anticipated Nvidia RTX 50 series GPUs have the potential to redefine the standards for GPU performance once they are released. While the extent of improvement offered by these GPUs remains uncertain, one thing is clear: Nvidia is determined to maintain its lead over AMD. Speculation is circulating that the next generation of Nvidia GPUs could be based on a consumer-focused version of ‘Blackwell’ architecture, further fueling excitement and anticipation. Now with Blackwell announced but for Generative AI, it’s only a matter of time before gaming GPUs show up.

Best GPU deals

---

Nvidia RTX 50 series release window prediction

The official release date for Nvidia’s RTX 5000 series GPUs remains undisclosed, while the RTX 4000 is still alive and kicking. However, the company has made comments regarding a potential successor to Ada, noting their intention to introduce a new graphics processing architecture every two years, as reported by Seeking Alpha. This implies that the lineup may debut sometime in 2024, although Nvidia’s plans are often subject to change.

And that has been the case for quite some time now. Looking back at the previous architecture release dates there is an obvious pattern.

• RTX 40 series – October 2022
• RTX 30 series – September 2020
• RTX 20 series – September 2018
• GTX 10 series – June 2016

So although the month may change, the year continues the same. So we should expect to see the RTX 50 series launch in the second half of 2024. Moore’s Law is Dead says that according to internal sources, Nvidia will be ready to launch Blackwell in Q4 2024 but it will depend on AMD’s RDNA 4. Either way, CES 2025 will be a key point to make on the efficiency of the generation.

There are plenty of reasons why it might be delayed. Firstly with the AI boom, Nvidia is actually at a shortage of enterprise GPUs. So it might be focusing the production of its silicon on that industry as it does have a much higher profit margin for the company and much larger volumes. As such the gaming market is less impactful and with the RTX 40 series not selling as well it might just show Nvidia it’s not as worth the resource investment.

Or, it could always be more development time. Allowing the previous generation to keep selling but improving what it has to offer with even potentially a refresh of the Ada range. As the Super lineup released we’ve seen the Ada range getting upgrades so might signal it’s coming to a close as well.

---

Nvidia RTX 50 series price prediction

The pricing of the upcoming RTX 50 series GPUs remains uncertain, but we can speculate based on a few educated guesses. Many gamers have expressed frustration over the high cost of Nvidia’s 40 series GPUs, with some believing that Nvidia took advantage of the Covid situation to increase prices, relying on gamers’ willingness to pay. However, it is yet to be seen if gamers will continue to accept these high prices.

Considering the backlash Nvidia faced with the release of the RTX 40 series GPUs, there is hope that the company will make the 50 series more affordable and accessible to a wider range of users.

There was a $100 difference between the RTX 3090 and the RTX 4090, coming in at $1,499 and $1,599 respectively. If we assume this upward trend to be the norm, we could see the flagship Nvidia RTX 50 series GPU start at $1,699. With the rising cost of materials, this would seem to be the most educated guess at this stage. Again, this is our speculation, nothing is confirmed. Otherwise looking at the price of the current lineup should give us a rough idea of what to expect in the next generation.

---

Nvidia RTX 50 series spec expectations

Again, there aren’t many specifications released for the Nvidia RTX 5000 series GPUs. So all we can do is say Nvidia will probably try to increase CUDA core count and develop their RT cores further for better ray tracing performance. The new GPUs are rumored to be based on some form of Blackwell architecture – with a number of SKUs currently up for evaluation. What we are certain of is that Nvidia will want to keep its lead over AMD, especially with their rumored RDNA 4 in the works.

We assume there will be advancements in DLSS before then, so maybe Nvidia will be updating the Tensor cores inside the 5000 series GPU for better AI processing. As is the way of the world right now. RGT also rumors that the generation will have a new SM structure and hyperspeed interlinks. As well as possibly denoising the RT pipeline.

Elsewhere, expect massive improvements in Ray Tracing performance – with a huge gulf between the 30 series and noticeable gains over the 40 series (more on this below). Again, pure speculation but typically we have seen this in prior launches. There may even be more of a move towards 8K gaming, now that the top-end 40 Series models seem to have nailed 4K Ultra both at native 4K and/or with DLSS. We will post any rumors or spec leaks we find down below.

What is MCM (Multi-Chip Module)?

Another part of the specs that could be a major upgrade for the RTX 50 series is the inclusion of Multi-Chip Module technology. This is a more advanced design of semiconductors, which as the name would imply, incorporates multiple small chips (known as chiplets) into one single bundle or chip package. Early reports point towards Nvidia exploring this technology and possibly taking advantage of it for their RTX 50 series GPUs.

The benefits of MCM include better power efficiency, power yields, and thermal design. We look forward to seeing what Nvidia’s new Blackwell architecture can deliver if combined with this MCM technology. This is just like what we’ve seen in the top-end RDNA 3 GPUs, with the RX 7900 XTX or XT splitting the die up. This separates the graphics compute die with a 5nm process and then the 6nm memory cache allowing AMD to also save costs on not the newest processes of TSMC that may cost more for the whole chip. Nvidia might want to achieve the same.

Nvidia Geforce RTX 50 series may use GDDR7 Memory

Following a report on MyDrivers, the Chinese tech site. A post states the RTX 5000 series GPUs may be using GDDR7 memory to gain a huge bandwidth boost over previous-generation GPUs. If this is true it will push the 50 series GPUS to new heights, if the 5000 series improves on cores, the process architecture, and the GDDR7 memory standard, then we are in for one formidable RTX 50 series family. It will probably push the 50 series GPUs closer to 70% performance gains at the high end of the spectrum.

As of now, no JEDEC standard has been released for GDDR7, which is still in the official development phase. However, Samsung revealed its efforts toward GDDR7 in late 2022, with the company now showing off its new VRAM modules. These could be included in the new generation of hardware. It has also been rumored that the memory in the new GPUs will utilize 2GB (16Gbit) modules first.

Ada-next to feature a 512-bit memory bus

If Nvidia were to utilize the new memory standard, it would likely have to increase the memory bus to get the most out of it. As kopite7kimi rumors, the flagship expected of the RTX 50 series, most likely the RTX 5090, will feature a 512-bit memory bus. That means if it uses 32 Gbps GDDR7 with that bus, it will have a 2,048 GB/s memory bandwidth. Over doubles what the RTX 4090 even has to offer and really solidifies it as a top-of-the-range GPU.

Kopite further gives information on what the Blackwell series might provide. Suggesting the data center GPUs will set 10 Texture Processing Clusters with 8 graphics processing clusters. The gaming SKUs may be 12GPS with 8TPS resulting in 192 SMs, which would result in a 24,576 CUDA core count. A massive overall boost already from the 16,384 CUDAs and 128 SMs in the 4090.

TSMC 4NP node and improved cache

Nvidia unveiled the Blackwell range of architecture but started off with its AI range of options first of course. But that still gives us an insight as to what the gaming GPUs will have to offer as well. As following the announcement more info from leaker kopite7kimi has emerged about what the flagship RTX 50 series processor will utilize.

Previously thought to have been using 3nm, it is now expected that GB202 will be on the same process as the AI processors. That is TSMCs 4NP node, a variation of the 5nm choice that is already being used in Ada and Hopper GPUs. This should allow for a 30% increase in transistor density for an improvement in efficiency and performance.

Also expected to be utilizing CuLitho technology that is aimed at improving the production and manufacturing of the AI Tensor and Gaming GPUs. There is also an expectation of big gains in L1 cache in the GB202 range compared to the AD102 and GA102 options.

---

Nvidia RTX 50 Series FAQs

Will there be a RTX 5000 series?

There will likely be an RTX 5000 series GPU, continuing the success of its graphics cards, Nvidia is likely going to continue producing them. The series is likely to continue the performance increases and potential price increases to improve the experience and revenue of the company. There are plenty of leaks and rumors already out about the Blackwell range so we can expect to see them soon.

When did the RTX 50 series come out?

The RTX 50 series has not come out yet, Nvidia is currently in the middle of producing the RTX 40 series. Which itself is a great improvement over the previous offering and so it will continue to release these cards for two years after the initial announcement. This is the usual cycle, so we expect the RTX 50 series to be seen somewhere at the end of 2024.

Will RTX 50 series GPUs support GDDR7 memory?

While unknown at this stage, rumors suggest that the RTX 50 series will introduce GDDR7 VRAM, which would improve greatly on GDDR6. The new spec has been shown off, and rumors suggest the new range will feature the technology inside of it.

Will RTX 50 series be more expensive than 40 series?

When the 40 series kicked off, the hefty price tag of the 4090 definitely caused a stir. Nvidia is working their way down to the entry-level cards to provide better value cards.

We expect much of the same from the 50 series. Backlash from overpriced GPUs could convince Nvidia to lower the cost just a bit though. Especially with AMD (and even Intel) looking to provide strong competition.

Nvidia RTX 50 series release date: Final word

That wraps up everything we know so far! The Nvidia RTX 50 series release date is definitely going to be a highly anticipated one. Even while we’re still working our way through the 40-series, people look towards the future of GPUs, especially if they feel a little let down by the current generation. We expect to see the full range of options coming in the next generation, with the likes of the RTX 5090, RTX 5080, RTX 5070, and RTX 5060 expected to pop up. But with recent releases, the date of when each will come is certainly unknown just yet.