features, models and which one you should choose

With the arrival of the new SUPER models the folks at NVIDIA have completed the revamp of their GeForce RTX 40 series. We do not expect any new product launches within this generation, so the models on the market right now are the ones that establish the definitive organization of the different ranges into which we can divide this family.

NVIDIA has broadly maintained the same organization we saw with the GeForce RTX 30 series, albeit due to the introduction of an important technology unique to the GeForce RTX 40, frame generation, the actual level of smoothness we can achieve with these new graphics cards is far beyond what we could expect with more powerful models of the previous generation.

Although the differentiation and positioning of each model in the GeForce RTX 40 series is relatively straightforward I have been able to confirm that. there is some confusion, and that this is largely given by the misinformation and malpractice. of some people who insist on saying things like the GeForce RTX 4070 is mid-range, or that the GeForce RTX 4060 is low-end. It doesn’t make any sense to call a product that cost more than 600 euros mid-range, and neither to call a 330 euros graphics card low-end.

To a large extent all this is a consequence of sensationalism, fanaticism and forced hatred to try to maximize visits, things that in the end do no good and that can generate false beliefs in the userleading them to make the wrong decision when they go to buy a new graphics card. The thing has reached such an extreme that recently a person told me that he had bought a GeForce RTX 4070 Ti to play in 1080p because he only had 12 GB of graphics memory, and that with a GeForce RTX 4060 he was not going to have enough.

Yes, it’s nonsense, and that’s why I recommend you be careful where you inform yourselves and what things you believe. This reality has also encouraged me to share with you this guide, in which we are going to discover the keys to the GeForce RXT 40 series, its most important features, the available models, their differences and which one you should choose depending on your needs.

Table of Contents

GeForce RTX 40 mid-range graphics cards

GeForce RTX 4060: ideal for 1080p

GeForce RTX 40

This graphics card is positioned as the entry model to the GeForce RTX 40 series, but don’t be confused, it is an entry-level model. mid-range model in its own rightand has a fully optimized configuration to provide an excellent experience in 1080p. It features all the exclusive technologies of the Ada Lovelace architecture, which means it is a state-of-the-art model, and its specification ratio makes it a more powerful model than it may appear at first glance.

The GeForce RTX 4060 can also offer a good experience in 1440p resolution, but the optimal level for this graphics card is 1080p. It stands out for having very low power consumptionand for being able to run ray-traced games smoothly in 1080p. An interesting option to mount a powerful, cool and affordable PC with which to play in this resolution and enjoy a perfect experience.

Specifications

  • AD107 graphics core in 5 nm.
  • 18.9 billion transistors.
  • Package area: 146 square mm.
  • Graphics Processing Clusters (GPCs): 3.
  • Texture Processing Clusters (TPCs): 12.
  • 3,072 shaders at 2.46 GHz.
  • 96 texturing units.
  • 48 raster units.
  • 96 fourth-generation tensor cores.
  • 24 third-generation RT cores.
  • 128-bit memory bus.
  • 8 GB of GDDR6 at 17 GHz with a bandwidth of 272 GB/s.
  • 15.11 TFLOPs of power at FP32.
  • 24 MB of L2 cache (453 GB/s effective).
  • TGP of 115 watts, average power consumption of 110 watts.
  • Uses PCIe Gen4 x8 standard.
  • Requires an additional 8-pin power connector.

GeForce RTX 4060 Keys.

  • It has all the advances of the GeForce RTX 40 series, such as “shader execution reordering”, third-generation RT cores and fourth-generation tensor cores.
  • Works with almost any type of power supply.
  • It performs beautifully at 1080p, and does well at 1440p.
  • Supports NVIDIA DLSS super resolution, frame generation and ray reconstruction, which improve image quality and can exponentially multiply the frame rate per second.
  • It is quite reasonably priced for what it offers (around €300).
  • At 1080p it gets to perform almost 20% better on average than the 12GB GeForce RTX 3060.

You should choose this model if your goal is to play in 1080p with all the guaranteesThe most advanced technologies in the industry and a high level of efficiency without having to make a large investment.

GeForce RTX 4060 Ti: better prepared for 1440p

GeForce RTX 40

This graphics card represents a considerable leap forward compared to the previous one, as it yields between 15% and 20% more at 1080pand like it retains all the key features and technologies of the Ada Lovelace architecture. The GeForce RTX 4060 Ti received some criticism because it came to market with 8GB of graphics memory, which prompted NVIDIA to later release a 16GB model, but the truth is that it delivers superb performance at 1080p and performs very well at 1440p.

Scenarios in which the GeForce RTX 4060 Ti may have problems due to lack of graphics memory. are minimal and are mainly limited to very specific situations at 1440p, so much so that they almost have to be forced to occur, and when this happens they are usually solved by resorting to the technologies clustered around NVIDIA DLSS. I highlight this because the misinformation that has been generated around graphics cards with 8GB of memory has been nonsense.

Specifications

  • AD106 graphics core in 5 nm.
  • 22.9 billion transistors.
  • Graphics Processing Clusters (GPCs): 3.
  • Texture Processing Clusters (TPCs): 17.
  • 4,352 shaders at 2,310 MHz-2,535 MHz, normal and turbo mode.
  • 136 texturing units.
  • 136 fourth-generation tensor cores.
  • 48 raster units.
  • 32 third-generation RT cores.
  • 128-bit bus.
  • 8 GB of GDDR6 at 18 GHz (288 GB/s bandwidth).
  • 22.46 TFLOPs of power at FP32.
  • 32 MB of L2 cache (554 GB/s effective).
  • Uses PCIe Gen4 x8 standard.
  • 160 watts TGP.

Keys to the GeForce RTX 4060 Ti

  • It has very low power consumption, so it doesn’t need a very powerful power supply.
  • It offers excellent performance at 1080p and performs well at 1440p.
  • Supports NVIDIA DLSS super resolution, frame generation and ray reconstruction, which improve image quality and can exponentially multiply the frame rate per second.
  • The 8GB version has come down quite a bit in price and is a more interesting option than the 16GB version for price-performance ratio.
  • It has all the advances of the GeForce RTX 40 series, such as “shader execution reordering”, third-generation RT cores and fourth-generation tensor cores.
  • It is very cool and offers an approximate 10% performance improvement at 1080p over the GeForce RTX 3060 Ti while consuming 40 to 50 watts less.

You should choose this graphics card if you want to enjoy a perfect 1080p experience with any current title, and it is also a good option to play with guarantees in 1440p without having to make a large investment.

Entry to high-end GeForce RTX 40 graphics cards

GeForce RTX 4070: ideal for 1440p, copes well in 4K

GeForce RTX 40

This graphics card is the entry-level model in NVIDIA’s high-end GeForce RTX 40 series, and from the very first moment it became one of the company’s most interesting options in terms of price-performance ratio, as it offers a similar raw power to the GeForce RTX 3080 at 1080p and 1440p.and lags slightly behind it in 4K.

The GeForce RTX 4070 Gives its best at 1440p resolution. At that level the experience is excellent even in demanding games, and with its 12 GB of graphics memory it performs flawlessly even in unrealistic situations where memory could become a limiting factor. It also stands out for its consumption, which is perfectly balanced for the range in which it is framed, and for the performance it offers.

Specifications

  • AD104 graphics core (Ada Lovelace architecture) in 5 nm.
  • 35.8 billion transistors.
  • Graphics Processing Clusters (GPCs): 4.
  • Texture Processing Clusters (TPCs): 23.
  • 5,888 shaders at 1,920 MHz-2,475 MHz, normal and turbo mode.
  • 184 texturing units.
  • 184 fourth-generation tensor cores.
  • 64 raster units.
  • 46 third-generation RT cores.
  • Power at FP32: 29.15 TFLOPs.
  • 192-bit bus.
  • 12 GB of GDDR6X memory at 21 GHz.
  • Bandwidth: 504.2 GB/s.
  • L2 cache: 36 MB.
  • TGP of 200 watts.

GeForce RTX 4070 Keys.

  • Excellent performance at 1440p and good at 4K in most cases.
  • Its 12 GB of graphics memory allows it to perform best at 1440p with maximum settings and active ray tracing.
  • It has all the advances of the GeForce RTX 40 series, such as shader execution reordering, third-generation RT cores and fourth-generation tensor cores.
  • Supports NVIDIA DLSS super resolution, frame generation and ray reconstruction, which improve image quality and can exponentially multiply the frame rate per second.
  • It consumes 120 watts less than the GeForce RTX 3080 and delivers similar performance. We don’t need a very powerful power supply to drive it.
  • It is available in fairly compact versions and models, so it fits in relatively small chassis.

At the time it was, as I said, a very interesting option for its price-performance value, but after the release of the GeForce RTX 4070 SUPER it has ceded the crown to this one. Therefore, I recommend that you pass on it and go straight for the SUPER model.for only 50-60 euros more you will enjoy a significant jump in performance.

GeForce RTX 4070 SUPER: more balanced

GeForce RTX 40

There is no better way to describe it, as it is an evolution of the GeForce RTX 4070 that features an AD104 GPU with a much higher shader count, and also has a higher amount of L2 cache, allowing it to deliver superior performance. Overall, the GeForce RTX 4070 SUPER has slightly lower performance than a GeForce RTX 4070 Ti, but cost about €200 less than that one, so it’s a better choice.

The GeForce RTX 4070 SUPER is very good at 1440p, and offers a fully optimal experience at that resolution even with demanding games. Its power consumption is very good for the performance it has, and it doesn’t need a very powerful power supply to work properly. In raw performance in 4K is almost at the level of a GeForce RTX 3080 Ti, and consumes 130 watts less.

Specifications

  • AD104 GPU manufactured in 5 nm process (TSMC).
  • 35.8 billion transistors.
  • Package area: 392 square mm.
  • Graphics Processing Clusters (GPCs): 5.
  • Texture Processing Clusters (TPCs): 28.
  • 7,168 shaders at 1,980 MHz-2,475MHz, normal and turbo mode.
  • 224 texturing units.
  • 80 raster units.
  • 224 fourth-generation tensor cores.
  • 56 third-generation RT cores.
  • 192-bit bus.
  • 12 GB of GDDR6X memory at 21 Gbps.
  • Uses PCIe Gen4 standard in x16 mode.
  • L2 cache: 48 MB.
  • Bandwidth of 504.2 GB/s.
  • Power at FP32: 35.48 TFLOPs.
  • TGP: 225 watts.

Keys to the GeForce RTX 4070 SUPER

  • Perfect performance for 1440p and very good in 4K.
  • Highly efficient, performs like a GeForce RTX 3090 at 1080p and 1440p, and consumes 130 watts less.
  • It has all the advances of the GeForce RTX 40 series, such as “shader execution reordering”, third-generation RT cores and fourth-generation tensor cores.
  • Supports NVIDIA DLSS super resolution, frame generation and ray reconstruction, which improve image quality and can exponentially multiply the frame rate per second.
  • Very attractive price for the performance it offers and what we can find in its range.

This should be our choice for play at 1440p fully optimally, even with ray tracing enabled, or at 4K with some compromises. in more demanding games. Its price-performance value is better than the GeForce RTX 4070so it’s a better buy overall.

High-end GeForce RTX 40 graphics cards

GeForce RTX 4070 Ti SUPER: Capable in 4K

This graphics card has displaced the GeForce RTX 4070 Ti, which is now discontinued and therefore will not appear in this guide. In terms of raw power the GeForce RTX 4070 Ti SUPER does not make a big difference compared to the GeForce RTX 4070 Ti, but it is configured with 16 GB of graphics memory, and this makes it a more versatile solution and better prepared for 4K.

Despite the slight performance boost it offers, and having more graphics memory, the GeForce RTX 4070 Ti SUPER has the same TGP as the GeForce RTX 4070 Ti, which means that its power consumption values have not increased, and that to run it we will not need a more powerful power supply. This confirms a good level of efficiency.

Specifications

  • AD103 GPU manufactured in 5 nm process (TSMC).
  • 45.9 billion transistors.
  • Package area: 379 square mm.
  • Graphics Processing Clusters (GPCs): 6.
  • Texture Processing Clusters (TPCs): 33.
  • 8,448 shaders at 2,340 MHz-2,610 MHz, normal and turbo mode.
  • 264 texturing units.
  • 96 raster units.
  • 264 fourth-generation tensor cores.
  • 66 third-generation RT cores.
  • 256-bit bus.
  • 16 GB of GDDR6X memory at 21 Gbps.
  • Uses PCIe Gen4 standard in x16 mode.
  • L2 cache: 49 MB.
  • 672 GB/s bandwidth.
  • Power at FP32: 45.02 TFLOPs.
  • TGP: 285 watts.

Keys to the GeForce RTX 4070 Ti SUPER

  • It’s great at 1440p and performs very well at 4K.
  • Good efficiency, it performs almost at the level of a GeForce RTX 3090 Ti in 4K and consumes 175 watts less.
  • It has all the advances of the GeForce RTX 40 series, such as “shader execution reordering”, third-generation RT cores and fourth-generation tensor cores.
  • Supports NVIDIA DLSS super resolution, frame generation and ray reconstruction, which improve image quality and can exponentially multiply the frame rate per second.
  • Performs up to 14% faster than the GeForce RTX 4070 Ti.
  • Its 16 GB of graphics memory allows it to work even with ray tracing in 4K, although at this level we will have to resort to DLSS technologies to achieve good fluidity.

This graphics card is an interesting option for those who want to play in 4K with guarantees without having to overcome the barrier of 1,000 euros.. Models are available from €879, and options with three-fan cooling systems from €919.

Top-of-the-line GeForce RTX 40 graphics cards.

GeForce RTX 4080 SUPER: 4K under control

Opening this range is the GeForce RTX 4080 SUPER, a model that has also displaced the GeForce RTX 4080, which is why I have not introduced the latter in the guide. The performance difference between the two is small, but the important thing is that NVIDIA has been able to reduce consumption the selling price.which makes this SUPER version more interesting than the standard version.

The GeForce RTX 4080 SUPER features excellent 4K performance even in demanding gamesand performs quite well even with ray tracing enabled at that resolution, although to achieve full fluidity with this technology it is necessary to pull the solutions that NVIDIA DLSS integrates.

Specifications

  • AD103 GPU manufactured in 5 nm process (TSMC).
  • 45.9 billion transistors.
  • Graphics Processing Clusters (GPCs): 7.
  • Texture Processing Clusters (TPCs): 40.
  • 10,240 shaders at 2,295 MHz-2,550 MHz, normal and turbo mode.
  • 320 texturing units.
  • 112 raster units.
  • 320 fourth-generation tensor cores.
  • 80 third-generation RT cores.
  • 256-bit bus.
  • 16 GB of GDDR6X memory at 23 Gbps.
  • Uses PCIe Gen4 standard in x16 mode.
  • L2 cache: 65 MB.
  • Bandwidth of 736 GB/s.
  • Power at FP32: 52 TFLOPs.
  • TGP: 320 watts.

Keys to the GeForce RTX 4080 SUPER

  • Fantastic 4K performance.
  • Performs 12% better than the GeForce RTX 3090 Ti at 4K while consuming 150 watts less.
  • It has all the advances of the GeForce RTX 40 series, such as “shader execution reordering”, third-generation RT cores and fourth-generation tensor cores.
  • Supports NVIDIA DLSS super resolution, frame generation and ray reconstruction, which improve image quality and can exponentially multiply the frame rate per second.
  • Outperforms the GeForce RTX 4080 by 3.3% on average at 4K, and consumes slightly less.
  • Having 16 GB of graphics memory, it has no problems in 4K with active ray tracing.
  • It performs almost the same as a Radeon RX 7900 XTX in rasterization, and beats it in ray tracing.

This is the best option to build a high performance PC designed for 4K gaming even with active ray tracing without blowing your budget out of the water.since we can buy a GeForce RTX 4080 SUPER from 1,119 euroswhich is 250 euros less than what the standard model cost at launch, and 720 euros less than what the cheapest GeForce RTX 4090 costs.

GeForce RTX 4090: maximum performance

The most powerful graphics card available in the mainstream consumer market, and also the most expensive. It far outperforms the GeForce RTX 4080 SUPER, and is capable of even at 4K in demanding titles such as Alan Wake 2. It undoubtedly offers the best possible experience at that resolution, and its overall power is impressive.

At 4K, the GeForce RTX 4090 performs at around 40% more on average than the GeForce RTX 3090 Ti and has an identical TGP, although its actual power consumption is lower, consuming on average 20 watts less. Compared to the GeForce RTX 4080 SUPER, the difference is as much as 25%, and it also has the advantage of 24 GB of graphics memory, making it a better choice than the GeForce RTX 3090 Ti. very capable even in demanding professional environments.

Specifications

  • AD102 GPU manufactured in 5 nm process (TSMC).
  • 76.3 billion transistors.
  • Graphics Processing Clusters (GPCs): 11.
  • Texture Processing Clusters (TPCs): 64.
  • 16,384 shaders at 2,235 MHz-2,520 MHz, normal and turbo mode.
  • 512 texturing units.
  • 176 raster units.
  • 512 fourth-generation tensor cores.
  • 128 third-generation RT cores.
  • 384-bit bus.
  • 24 GB of GDDR6X memory at 21 Gbps.
  • Uses PCIe Gen4 standard in x16 mode.
  • L2 cache: 72 MB.
  • Bandwidth of 1.008 GB/s.
  • Power at FP32: 82.58 TFLOPs.
  • TGP: 450 watts.

GeForce RTX 4090 Keys.

  • Fully optimized 4K performance.
  • Outperforms the GeForce RTX 3090 Ti by 40% and consumes less power than the GeForce RTX 3090 Ti.
  • It is very cool, its temperature values remain at optimal levels even when it exceeds 400 watts of power consumption.
  • It has all the advances of the GeForce RTX 40 series, such as “shader execution reordering”, third-generation RT cores and fourth-generation tensor cores.
  • Supports NVIDIA DLSS super resolution, frame generation and ray reconstruction, which improve image quality and can exponentially multiply the frame rate per second.
  • It is capable of delivering good performance even with ray tracing without resorting to DLSS technologies. With DLSS technologies active, it offers full fluidity even with ray tracing (full ray tracing).
  • It has fantastic performance even when working with professional applications and demanding projects.

You should go with the GeForce RTX 4090 if you want to enjoy the ultimate gaming experience under 4K resolutionand it is also a good investment if you are going to use them. for gaming and for work. The latter would, in fact, be the ideal scenario, because it represents a significant investment due to its price of almost €1,900.

Click to rate this entry!
(Votes: 0 Average: 0)
Share!

Leave a Comment