TOP 7. Best video cards for games. June 2020. Rating!

Hello, everyone! Whenever you need to replace or buy a video card, it is an important step that may improve or decrease the the level of gaming performance of your computer. To make the best choice, you need to take into account many factors, such as frequency of operation, memory size, wire width, etc. In this video I’m going to talk about the best video cards for games at the moment to simplify this process for you.

Nvidia GeForce GTX 1650 Super

  • Architecture: Turing
  • Process: 12 nm
  • Number of universal processors: 1280
  • Number of texture blocks: 80
  • Number of rasterization blocks: 32
  • Game clock frequency: 1530 MHz
  • Turbo clock speed: 1725 MHz
  • Video memory type: GDDR6
  • Video memory: 4 GB
  • Power Consumption: 100 W
  • Ray tracing: no

At the end of November last year NVIDIA introduced the GeForce GTX 1650 Super graphics card. Compared to the “regular” GTX 1650, this has a more powerful TU116 GPU processor which works in the older GTX 1660 and GTX 1660 Ti models. The chip area also increased to 284 mm sq., and the number of transistors is up to 6.6 billion. The base clock frequency has also grown, it is 1530 MHz. The memory interface remained 128 bit, but now it uses GDDR6, so that the memory bandwidth has increased to 192 GB / s. The presence of a larger and faster GPU led to an increase in power consumption: declared by the manufacturer, its level is 100 watts now.

While creating this video card Nvidia used a stripped down version of the GPU from the GTX 1660. As a result, the more powerful processor provided a noticeable performance increase when compared with the previous version of this graphics accelerator.

The GTX 1650 Super is 35% faster than the Radeon RX 570. Compiaring to the Radeon RX 580 which has 8 GB of memory, the GTX 1650 Super shows roughly the same performance in many games.

In short, the GeForce GTX 1650 Super is a budget graphics card that will allow you to feel comfortable in almost any modern games at medium and high graphics settings.

Nvidia GeForce GTX 1660 Super

  • Architecture: Turing
  • Process: 12 nm
  • Number of universal processors: 1408
  • Number of texture blocks: 88
  • Number of rasterization blocks: 48
  • Game clock frequency: 1530 MHz
  • Turbo clock speed: 1785 MHz
  • Video memory type: GDDR6
  • Video memory: 6 GB
  • Power Consumption: 125 W
  • Ray tracing: no

NVIDIA, continuing to compete with AMD Radeon graphics cards, has released one of the best graphics cards designed for gaming on monitors with a resolution of 2560×1440 and 1920×1080. This card is the GeForce GTX 1660 SUPER.

In general this is the same GTX 1660 with the TU116-300 graphic core, which contains 1408 CUDA cores, 22 SM blocks, 48 ​​raster and 88 texture blocks. The frequencies also have not changed and are 1530 MHz in the base, and 1785 MHz in the turbo mode. The performance has grown significantly thanks to the transition from the GDDR5 to the modern GDDR6 memory. The effective clock speed also increased from 8 to 14 GHz, and the bandwidth increased from 192 to 336 GB / s. According to these indicators, the 1660 SUPER even outperformed the GTX 1660 Ti, with its clock frequency of 12 GHz and a bandwidth of 288 GB / s.

The NVIDIA developers call this model the best upgrade option with the GTX 1060, and promise a performance increase of 30 to 50%.

For quite some time, the NVIDIA’s most powerful graphics accelerator without RTX technology was the GTX 1660 Ti. Now the cheaper 1660 SUPER with the same level of performance took its place. In this regard, it also overtakes the Radeon RX 590 which is currently outdated and is not able to fully compete with new models of video cards.

In short, if you want to play in 1080p or 1440p at high graphics settings and plan to build a powerful gaming computer without overpayment, then the GeForce GTX 1660 SUPER is your choice.

AMD Radeon RX 5600 XT

  • Architecture: Navi RDNA
  • Process: 7 nm
  • Number of universal processors: 2304
  • Number of texture blocks: 144
  • Number of rasterization blocks: 64
  • Game clock frequency: 1375 MHz
  • Turbo clock speed: 1560 MHz
  • Video memory type: GDDR6
  • Video memory: 6 GB
  • Power Consumption: 550 W
  • Ray tracing: no

Next on our list is the AMD Radeon RX 5600 XT which is equipped with a Navi 10 GPU running on the RDNA architecture. This processor was previously the basis for the Radeon RX 5700 XT and the RX 5700. In other words, instead of creating a new chip, AMD decided to use the proven option and simply cut back on the capabilities of the older processor. As a result, the budget Navi 10 received the 2304 stream processors, 64 rasterization units and 144 texture units. Compared to the older model, the memory bus was cut to 192 bits by the developers, and the memory size was reduced to 6 GB.

But even this stripped-down version with less memory, a 192-bit bus and a lower core frequency shows higher performance than the video card that we just looked at – the GTX 1660 Super. However, it affects the cost of these products. If we talk about the overclocked version of the RX 5600 XT, then it can get about 10% increase in speed. But the main surprise of this video card was its low power consumption, as well as its very quiet operation.

In general, according to a combination of characteristics important for the consumer, this is a very successful graphics accelerator in its price segment.

Nvidia GeForce RTX 2060 Super

  • Architecture: Turing
  • Process: 12 nm
  • Number of universal processors: 2176
  • Number of texture blocks: 136
  • Number of rasterization blocks: 64
  • Game clock frequency: 1470 MHz
  • Turbo clock speed: 1650 MHz
  • Video memory type: GDDR6
  • Video memory: 8 GB
  • Power Consumption: 175 W
  • Ray tracing: yes

The Nvidia GeForce RTX 2060 Super externally practically does not differ from its predecessor – the RTX 2060. But the filling of these video cards is different. The developers equipped the super version of the RTX 2060 with 8 GB of memory, and placed 8 memory chips on the carrier board. Thanks to which the video processor has more room for maneuvering. The GPU remains the same, but now it has 2176 CUDA cores, other features have also increased. Along with the increase in productivity, the level of energy consumption also increased, the nominal indicator of which is 175 watts.

It is worth noting that the new video adapter is about 14% faster than its predecessor, and in heavy games the difference reaches 19%. The result is expected, because in addition to increasing the blocks in the GPU, the Super variant also received a 256-bit bus, as well as 2 additional GB of memory. As a result, in terms of performance and basic characteristics, this video card can easily compete with the GeForce RTX 2070. If you compare this model with its AMD competitors, the RTX 2060 Super will be somewhere between the RX 5700 and the RX 5700 XT.

As a result, the GeForce RTX 2060 Super is a very quiet and “cold” video card with sufficient performance for a comfortable gaming with a resolution of 1440p and 1080p at high graphics settings.

Don’t forget that you can find all the current prices for these graphics accelerators just by clicking on the links that I left for you in the description below the video.

AMD Radeon RX 5700 8GB

  • Architecture: Navi RDNA
  • Process: 7 nm
  • Number of universal processors: 2304
  • Number of texture blocks: 144
  • Number of rasterization blocks: 64
  • Game clock frequency: 1465 MHz
  • Turbo clock speed: 1725 MHz
  • Video memory type: GDDR6
  • Video memory: 8 GB
  • Power Consumption: 600 W
  • Ray tracingRay tracing: no

Next in line is an interesting model from AMD. This is the Radeon RX 5700. The RDNA architecture, as practice has shown, met the expectations of users. The advanced technical process and completely redesigned logic of the Navi 10 GPU allowed AMD to compete with NVIDIA accelerators in the mid-high price range, for example with the GeForce RTX 2060 and the RTX 2070. If you look at the aggregate data in gaming tests, the RX 5700 is ahead of the RTX 2060 by 6 –10%, depending on the resolution of the display. The RX 5700 XT goes almost on a par with the RTX 2070. However, NVIDIA counterattacked by releasing the Super models, three of which we have already examined in this review. Although the Navi graphics accelerators outperform the competitors from the “green camp” in some parameters, the Super video cards are usually ahead in tests.

Nevertheless, the RX 5700 has many advantages, among which are sufficient performance for games in 1080p resolution with high quality settings, 8 GB of video memory, excellent power-to-performance ratio, low noise level at low loads, as well as an adequate price.

Nvidia GeForce RTX 2070 Super

  • Architecture: Turing
  • Process: 12 nm
  • Number of universal processors: 2560
  • Number of texture blocks: 160
  • Number of rasterization blocks: 64
  • Game clock frequency: 1605 MHz
  • Turbo clock speed: 1770 MHz
  • Video memory type: GDDR6
  • Video memory: 8 GB
  • Power Consumption: 215 W
  • Ray tracing: yes

The GeForce RTX 2070 Super, as the name implies, is an accelerated modification of the RTX 2070. Nevertheless, based on the parameters and the level of performance, it is more similar to the older model in this line – the GeForce RTX 2080. The main innovation of the video card is the use of a more powerful TU104 graphics processor. As a result, compared with the base model, the number of CUDA computers has increased by 11%, there are also more texture units – now there are not 144, but 160. The hardware RT computers and tensor kernels have proportionally increased to 40 and 320, respectively. But the memory subsystem remains the same. The video card is equipped with 8 GB of high-speed GDDR6 memory, which runs at 14,000 MHz. Thanks to this, the 256-bit bus provides good bandwidth – 448 GB / s.

The presence of a large cooling system allows you to use the device during game battles in a passive mode, with a minimum load and noise level. The video card initially has factory overclocking, which can be further increased if desired. In general, the GeForce RTX 2070 Super is perfect for monitors of 1920 × 1080 with a frequency of 144 Hz. At the same time it copes with dignity even with heavy games in 1440p mode and high graphics settings.

Nvidia GeForce RTX 2080 Ti

  • Architecture: Turing
  • Process: 12 nm
  • Number of universal processors: 4352
  • Number of texture blocks: 272
  • Number of rasterization blocks: 88
  • Game clock frequency: 1350 MHz
  • Turbo clock speed: 1545 MHz
  • Video memory type: GDDR6
  • Video memory: 11 GB
  • Power Consumption: 250 W
  • Ray tracing: yes

The GeForce RTX 2080 Ti is currently the most powerful gaming graphics card with unprecedented levels of performance and support for all new NVIDIA technologies. It shows a significant increase in performance when compared to the models of the previous generation. The new flagship in heavy graphics tasks is ahead its predecessor – the GTX 1080 Ti by about 37%! Compared to the RTX 2080, the performance in 4K resolution has grown by 25-30%. Nevertheless, we have to say that in lower resolutions the difference is not so noticeable.

The similar power of this video card is provided by the new TU102 graphic processor, which has a record number of computing units. The work of shaders has also improved, which will provide 1080 Ti with additional benefits in the future. The new architecture also should be mentioned, as well as the availability of hardware support for ray tracing, and deep learning technology (for example, the DLSS smoothing). It is logical that this GPU has a fairly high heat dissipation and power consumption.

Unfortunately, despite all the advantages of the RTX 2080 Ti it has a very high price for the video card segment. However, this does not stop wealthy gamers who want to personally evaluate the new capabilities of the Turing architecture and get a game accelerator with maximum performance.

Rate article
Add a comment