What is known about NVIDIA GeForce RTX 5000 series graphics cards

As the current generation of NVIDIA GeForce RTX 4000-series graphics cards cease to impress with their performance, the eyes of enthusiasts and professionals are gradually turning to the future lineup. GeForce RTX 5000. Despite the fact that there is no official information from NVIDIA on this matter yet, the earth is full of rumors and assumptions about the characteristics and capabilities of the next generation of graphics accelerators. As a result of forecasts and leaks, there are already enough to at least try to form a picture of what to expect from the new video cards.

When will the GeForce RTX 5000 come out?

Initially GeForce RTX 5000 series graphics cards release was expected no earlier than mid-2025. However, recently information about the possibility of an earlier release has been increasingly slipping out. The key factor that could influence the reduction of the preparation time for the new line is said to be the increasing competition from AMD, which in recent years has demonstrated impressive progress in both the desktop and mobile GPU segments.

The success of the RDNA architecture and its iterations is forcing NVIDIA to be more aggressive in developing and releasing new products. The new iGPU alone is worth it Radeon 890Mwhich turned out to be more powerful than the laptop version of the RTX 3050. It’s scary to imagine what the “reds” will offer in the discrete video card segment.

If the rumors are true and NVIDIA is indeed preparing an accelerated release of the RTX 5000, this could mean the following:

• The company is confident of significant progress in its technologies and is ready to introduce a significant increase in productivity earlier than originally planned.

• NVIDIA sees the potential of future AMD products (presumably the Radeon RX 8000 series) as a serious threat to its dominance in the high-performance GPU market.

• A company may make some compromises in terms of energy efficiency or technological sophistication in order to get to market earlier.

But even in this scenario, we shouldn't expect the release of new video cards before the end of 2024 – beginning of 2025. The development and production of such a complex product takes a significant amount of time, and NVIDIA simply cannot afford to release a raw and unusable product.

Blackwell Architecture: Evolution or Revolution?

According to available information, GeForce RTX 5000 series will be based on a new architecture codenamed Blackwell. This is a continuation of the line of NVIDIA architectures named after outstanding scientists – after Turing, Ampere and Ada Lovelace, it is time to immortalize the name of David Blackwell, a famous American mathematician and statistician.

Although the details Blackwell architecture While still under wraps, it's safe to assume that it will build on the ideas laid out in Ada Lovelace, with a focus on further improving performance in ray-traced scenarios and using AI to improve image quality.

Blackwell is expected to bring the following improvements:

• Increased number of CUDA cores, RT cores (for ray tracing) and Tensor cores.

• Increased clock speeds.

• Improved energy efficiency.

• Using a new technological process.

For a long time, rumors about the possible use in the RTX 5000 were of particular interest multi-chiplet design. However, according to the latest information, the top-end GB102 graphics chip, which will form the basis of the next-generation video cards, will have a monolithic structure. It is rumored to operate at a base frequency of almost 2.9 GHz, which is a significant step up from previous generations.

RTX 5000 Performance: An Expected Quantum Leap

While there is no concrete performance data for the RTX 5000 yet, insiders are expecting a significant boost over the current generation. The most interesting information concerns the flagship model – the supposed GeForce RTX 5090According to rumors, this video card may be approximately 1.7-2 times more powerful than the current one. RTX 4090.

This performance gain can be achieved through the following improvements:

• Increased core count: 192 streaming multiprocessors and 24,576 CUDA cores expected.

• Increase memory bandwidth.

• Increased cache size.

• Increased clock speeds: base frequency will be 2.9 GHz.

• 12 GPC (Graphics Processing Clusters) clusters with 8 TPC (Texture Processing Clusters) each.

Energy consumption and efficiency: balancing power and ecology

One of the key issues related to RTX 5000 releaseis the energy consumption. Current generation NVIDIA flagship graphics cards has already come very close to the limits of standard power supplies and cooling systems in personal computers. The RTX 4090, for example, has a declared power consumption 450 W.

Sources say NVIDIA is looking to make the RTX 5000 series more power efficient. This is critical not only for usability and compatibility with existing PC components, but also in the context of the growing focus on greener high-tech products.

Achieving increased energy efficiency with significant productivity gains is a complex task that will require innovative solutions:

• Using a more advanced process technology to produce chips, perhaps 4nm or even 3nm.

• Optimization of architecture for more efficient energy use.

• Implementation of new chip-level power management technologies.

• Possible use of new materials in GPU design for better heat dissipation.

If NVIDIA can save power consumption of the flagship RTX 5090 model at a level close to the RTX 4090, with the declared increase in performance, this will be considered a great success. However, according to other data, we will see a further increase in the TDP of high-end video cards, and in the case of the RTX 5090, this value may be 500-600 W.

Because of this, NVIDIA is rumored to be considering using a new 16-pin power connector standard PCIe 6.0in order to thereby improve the reliability of the connection after problems with 12VHPWR connectors on RTX 4090.

Memory and Bandwidth: New Horizons

The significant increase in memory bandwidth mentioned by various sources (up to 52% compared to the RTX 4090) points to the possible use of new types of memory:

A memory increase is also likely, especially for flagship models. If The RTX 4090 comes with 24GB of GDDR6XThat RTX 5090 could get 28GB or even 32GB. This is especially important for professional applications, working with high resolutions and complex textures in games.

In addition, with each generation NVIDIA introduces new technologies that improve image quality and performance. Therefore, with a significant hardware upgrade, we can expect further development of solutions like DLSS. In particular, it is possible that DLSS 4.0 or similar technology will appear with even more impressive results.

And given the growing interest in AI technologies, we can expect NVIDIA to likely beef up its GPUs' capabilities in this area. This could include increasing the number and efficiency of tensor cores, as well as new software tools for developers.

How much will the RTX 5090 graphics card cost?

Now, regarding the price, given the expected significant performance gains and advanced technologies, we can expect the prices to be top models RTX 5000 will be sky-high. Various sources do not cite exact figures, but assure that one should focus on approximately 2000-2500 dollars.

This pricing policy may cause a mixed reaction among consumers, especially given the overall economic situation. However, NVIDIA will likely position the top RTX 5000 models as premium products for enthusiasts and professionals willing to pay for maximum performance.

And to satisfy less demanding users, even before the release of the RTX 5090, NVIDIA plans to launch RTX 5080 sales. It is rumored to have performance comparable to RTX 4090D, which is sold in Chinabut its price will be much more humane. Insiders claim that the 80-ka can be purchased for a little over $1,000. This is a much more interesting offer, and, obviously, it has every chance of becoming a bestseller.