Nvidia releases its RTX 30XX graphics cards. Several models are available for all budgets. Variants, prices, technical specifications and comparison of architectures, here is everything you need to know about Nvidia graphics cards.
Graphics cards Nvidia are on the RTX 30XX generation. It uses the Ampère architecture and wants to democratize the use of ray-tracing as well as technologies such as DLSS. The first models were announced in September 2020 and since then they have been joined by new cards as well as variations.
In total, we therefore have eight cards available. Here is the list, from most powerful to least powerful:
- GeForce RTX 3090
- GeForce RTX 3080 Ti
- GeForce RTX 3080
- GeForce RTX 3070 Ti
- GeForce RTX 3070
- GeForce RTX 3060 Ti
- GeForce RTX 3060
- GeForce RTX 3050
If you are lost, don’t panic. We will clarify all that in this file. What is the best card for your uses? How much are they sold for? We tell you everything!
📆 Are Nvidia’s Geforce RTX 30XX available?
The Geforce RTX 30XX are theoretically available, except for the latest, the RTX 3050. This was indeed presented at CES but will not be available until the end of January 2022.
However, the manufacturer’s graphics cards are very difficult to find. Since the release of the first models at the end of 2020, stock shortages have continued to annoy potential buyers. The shortage of semiconductors is obviously to be pointed out, but also the mining farms which get their hands on thousands of cards to the detriment of the players.
At the beginning of 2022, the situation is still not resolved and you still have to either be patient or break the bank to buy one at an indecent price. If Nvidia promises a reduction in the shortage this year, we are still not out of the woods. Additionally, Ethereum is expected to replace its Proof-of-Work (PoW) algorithm with a Proof-of-Stake (PoS) algorithm soon. Therefore, Nvidia graphics cards will be useless for mining cryptocurrency.
The proof is the problems encountered by the manufacturer to produce its new models. While the RTX 3090 Ti was originally due to be released in January 2022, this one is hiding in silence while the graphics card has still not shown the tip of its nose. The fault with a lack of components which prevents to manufacture the GPU in sufficient quantities.
💸 How much do Nvidia graphics cards cost?
The firm has made a remarkable effort in the field of prices knowing that its new cards make a huge leap forward in terms of performance. Here are the official prices of the RTX 30XX:
- GeForce RTX 3050: 250 euros
- GeForce RTX 3060: 325 euros
- GeForce RTX 3070: 519 euros
- GeForce RTX 3080: 719 euros
- GeForce RTX 3090: 1549 euros
As you will have noticed, the price has noticeably since the GTX. On average, it takes at least 100 euros more per equivalent model. An increase which Nvidia does not hesitate to praise the merits to its shareholders, arguing that players now spend 300 dollars more than before to buy their graphics card.
Indeed, at the time of writing these lines, the cards officially on the market are chronically out of stock. Their price sometimes reaches peaks on eBay, when manufacturers like MSI do not inflate the prices themselves. The competitive environment has also become more complex in the graphics chip market. AMD has become very aggressive with its new RDNA architecture.
Nvidia therefore offers its cards at fair prices… in theory. Today, it is impossible to find brand products at the displayed price. The entry-level is also not spared, while the RTX 3050 is displayed at prices up to 300€ higher than that recommended.
However, the situation is gradually improving. After several months of constant decline, the RTX 3000s are showing in March 2022 at the lowest price since January 2021. On eBay, prices have fallen by 25% in just 3 months, gradually approaching the MSRP. In May 2022, Nvidia cards are displayed at only 10% above the recommended price.
⚙️ What is the technical sheet of the RTX 3060, 3070, 3080 and 3090?
Nvidia offers real developments on paper. The GeForce RTX 3050, 3060, 3070, 3080 and 3090 are particularly enticing. We are indeed on an engraving almost twice as fine as the previous one, with a new architecture which explodes the memory bandwidth, and particularly muscular VRAM options:
| NVIDIA GeForce RTX 3070 | NVIDIA GeForce RTX 3080 | NVIDIA GeForce RTX 3090 | |
|---|---|---|---|
| GPU (architecture) | Ampere GA104-300 | Ampere GA102-200 | Ampere GA102-300 |
| Fineness of engraving | 8nm | 8nm | 8nm |
| CUDA Hearts | 5888 | 8704 | 10496 |
| Shader Flops | 20 TFLOPS | 30 TFLOPS | 36 TFLOPS |
| RT Flops | 40 TFLOPS | 58 TFLOPS | 69 TFLOPS |
| Tensor Flops | 163 TOPs | 238 TOPs | 285 TOPs |
| GDDR memory | 8 GB GDDR6 | 10 GB GDDR6X | 24G0 GDDR6X |
| Memory bus | 256-bit | 320-bit | 384-bit |
| Memory speed | 16Gbps | 19Gbps | 19.5Gbps |
| Memory bandwidth | 512 GB/s | 760 GB/s | 936 GB/s |
| Price | 519 euros | 719 euros | 1549 euros |
| Release date | October 2020 | September 17, 2020 | September 24, 2020 |
RTX 3090: the most powerful of all
It’s the ultimate from the Serie. The RTX 3090 clearly gives the impression of replacing the Titan range, as its performance is above the rest. It is equipped with a 24 GB GDDR6X VRAM, it is more than twice the 11 GB of the RTX 2080 Ti. As for raw power, it delivers 36 TFLPOPS for Shader Cores, 69 TFLOPS for RT Cores, and 385 TFLOPS for Tensor Cores. With this reference, Nvidia is aiming for 8K at 60 FPS and has also made demos in this direction during its conference.
In addition to video games, artificial intelligence and other advanced applications, this new card would also be particularly good at cracking passwords.
RTX 3080: twice as powerful as the RTX 2080 Ti
It is the choice of balance. Much more suitable for the vast majority of players than the RTX 3090 which seems ahead of its generation for a price that is cooling. Nvidia announces twice the performance compared to the RTX 2080 Ti, the most powerful of the Turing generation. At least, for the general public.
The RTX 3080 packs 10GB of GDDR6X RAM and delivers a mighty raw 30 TFLOPS for Shader Cores, 58 TFLOPS for RT Cores, and 238 TFLOPS for Tensor Cores. You have there a graphics card cut for 4K at 60 FPS without a hitch, with ray tracing activated please.
RTX 3070: the best value for money with performance on the order of the 2080 Ti
Here is the card that replaces the RTX 2080 Ti in terms of quality / price ratio. It is indeed more efficient while being offered at almost 2/3 of the price of the RTX 2080 Ti. An official benchmark confirms performance 8% above, on average, that of the 2080 Ti.
It is therefore an interesting proposition for those looking for the best value for money. The RTX 3070 comes with 8 GB of standard GDDR6 memory and delivers raw power of 20 TLOPS for Shader Cores, 40 TFLOPS for RT Cores and 163 TFFLOPS for Tensor Cores.
RTX 3060: the best value for money with performance on the order of the 2080 Ti
The RTX dedicated to those who want to play without breaking the bank is equipped with 3584 CUDA cores, 12 GB of memory on a 192-bit bus and a base frequency of 1.32 GHz which can go up to 1.78 GHz in boost. It has a TDP of only 170W, 10W more than the RTX 2060. We advise you to use a power supply of at least 550W to install it on your PC. The RTX 3060 is compatible with the DLSS, RTX, Reflex and other NVIDIA proprietary technologies. As far as connectors are concerned, we find without surprise HDMI 2.1 and DisplayPort 1.4a ports.
RTX 3050: Ray-Tracing in Full HD at 60 FPS for less than €250
This entry-level model is intended above all for modest budgets. Available from €250, this GPU can run the latest games in Full HD and 60 FPS or more. Also based on the Ampere architecture and 3rd generation Tensor cores, it features 8 GB of GDDR6 RAM and is also compatible with the latest Nvidia technologies, namely Ray Tracing and DLSS 2.0. The RTX 3050 will be available from January 27, 2022 starting at €250.
What is the performance of the mobile versions?
Nvidia markets versions of its cards for laptops. Obviously, they bring more performance than the portable RTX 20XX models. However, they remain below their desktop model, which is logical. For example, the RTX 3080 is much less capable than its classic iteration.
There is really nothing surprising here. These cards have very high power dissipation, and because of this, manufacturers have no choice but to severely limit performance to maintain a TDP below 150 W. However, it is possible that optimized computers can reduce the gap.
However, it is worth noting that on newer laptops, some less high-end models can perform very well. Indeed, thanks to the manufacturers’ optimization efforts, the latter can obtain better benchmark results than their supposedly more powerful counterparts. So be sure to take a good look at the different tests before making your choice.
What is the difference between Ampere and Turing architecture?
Ampere
With the Ampère architecture, Nvidia switches to 8nm engraving. The name of the Ampère microarchitecture derives its name from the mathematician and physicist André-Marie Ampère to whom we owe many discoveries in electronics.
The architecture’s new RT Cores deliver 2X the power of the previous generation for ray tracing processing. Ditto for third-generation Tensor cores that make it faster and more efficient to perform AI-based tasks, like NVIDIA DLSS. Upscaling technology recently passed the 150 compatible game mark, making it increasingly attractive to gamers. The new architecture pushes performance boundaries in all areas.
Turing
The Turing architecture was introduced in August 2018. This architecture is the first on a consumer product to enable Ray Tracing. It offers, in addition to cores dedicated to Ray Tracing, cores dedicated to artificial intelligence (Tensor cores). Note that a variant of the Turing architecture without Ray Tracing exists. The Turing architecture is engraved in 12nm using the FinFET (TSMC) process. The highest-end GPU, the TU 102, features 18.6 billion transistors. The Turing architecture is also the first to support GDDR6 memory.
The CUDA cores have been revised from release that the Turing architecture allows, according to Nvidia, performance gains between 30 and 50% on existing games compared to the previous generation. On the Ray Tracing side, the performance would have been multiplied by four compared to what the highest-end card under Pascal architecture allowed. As for the cores dedicated to artificial intelligence, they are most of the time used to fill in missing parts of the image (a technique called de-noising).
It is these cores that are used, for example, when the definition exceeds the original one. Also, these cores do not do artificial intelligence directly. The models are created on supercomputers which define by machine learning the most appropriate method for the AI cores (Tensor Cores) of the cards.
