Nvidia rtx 2060 super review in Spanish (full analysis)

Nvidia RTX 2060 Super and the new evolution of Nvidia's RTX graphics is here. And it is that practically the three models to launch will rise one step compared to the previous RTX, resulting in a performance of only 1% lower in this RTX 2060 Super compared to the RTX 2070. We are looking forward to verifying this new performance, but with 8 GB GDDR6, a 256 bit bus and 2176 CUDA Cores we can intuit that the performance will be on par.

As it could not be otherwise, Tensor and RT cores have been included to take the performance of new and next generation games to a new level, in 1440p with Ray Tracing and DLSS giving the maximum of them. Do you want to see this GPU in action? Well, let's go there!

But first, we must thank Nvidia for the trust placed in us by giving us these new GPUs to do our analysis.

Nvidia GeForce RTX 2060 Super technical features

Unboxing

Well, we started one more review with this new generation of graphics cards, or rather, this new update to compete with the new GPUs from AMD. And as is tradition, we must start with the Unboxing of Nvidia GeForce RTX 2060 Super.

And the truth is that precisely the presentation is not a novelty, since like the previous GPUs, we find a rigid and very thick cardboard box with a print practically identical to normal RTX, based on the color green and in gray colors in those lines that make the view dizzy.

On the sides we find the typical information about the product, with the RTX technology clearly visible. The opening system will also be the same, upwards with the graphics card placed vertically in a high-density foam mold.

Inside, we only have the graphics card, wrapped in a plastic sheet and with no other accessories than the user manual and the warranty card.

Exterior design

And here we have the final product, a graphics card that maintains its spectacular and elegant heatsink. In fact, the only noticeable difference with the RTX 2060 is that it now features the "Super" badge on it in the center. Both the thickness and the colors chosen are the same and therefore do not stop being the most accurate. What does change, is the central area of ​​its distinctive, now being completely chrome, and giving it an even more Premium appearance. Far away are those turbine sinks that gave such low performance in continuous stress processes.

The main block or housing of this heatsink is made of aluminum with a matte finish of a certain roughness and the natural color of this metal. In addition, we note that it is made entirely in a single helmet to provide more rigidity to the set. The dual-fan configuration is maintained, as is the entire range of new-generation GPUs and perfectly on par with custom models.

This cooling configuration has two 85mm diameter fans with a total of 13 flat fins each with a dual-axis core that allows maximum durability and a very quiet system. But just like in the previous models, these fans will always be working. So we don't have the typical systems that only activate the fans when we go over a certain temperature.

If we turn it is Nvidia GeForce RTX 2060 Super to see the upper area, we find a nice backplate also built in quite thick aluminum with the characteristic square grooves that, in addition to aesthetics, have also been designed to support greater weight and provide greater dissipation surface of heat. It has anti-scratch protection, and is fixed to the upper area by means of those small screws that we see at each end.

The function of the other screws is to attach the internal aluminum heatsink itself, the fans and also the heatsink to the graphics processor. Everything very visible as always in this configuration and ready to be opened or at least try.

The side area maintains that elegant aesthetic provided with steps in which a large “GEFORCE RTX” sign is displayed, which will light up green when the GPU is installed, it is time for Nvidia to join the RGB club, although green is its corporate color. Although the truth is that at first glance it seems that it is simply silver. The internal aluminum heatsink with a high density of finned and painted in black is very visible .

Ports and power connections

If we place ourselves in the back, we will be able to see how the port panel of this Nvidia GeForce RTX 2060 Super, which does not come with too many new features compared to the normal model. Obviously, it is a double slot configuration with the following ports:

• 1x HDMI 2.0b2x DisplayPort 1.41x DVI-DL1x USB Type-C

So it supports a total of four monitors and a wide variety, available from up to three different connectors. Nvidia continues betting on this inferior model for a Digital connector for older monitors, presumably at Full HD resolution.

For its part, the HDMI connector will support a maximum resolution of 4K (3840x2160p) at 60 FPS, while the double Display Port supports resolutions of up to 8K for high-performance monitors for design. In any case, with this card we will be able to move comfortably in 1080p or 1440p resolutions for games.

As for the power connector, we have it located again in the front area of ​​the card in a very original way and more hidden than the other models. And again the configuration is repeated, being 8-pin EPS, we must remember that the TDP of this Nvidia GeForce RTX 2060 Super amounts to 175 W, compared to 160 that has the normal version. Anyway, Nvidia considers that it will be enough for the model, even with overclocking? We will see later, but taking into account that the normal RTX 2070 also has an 8-pin connector and the same TDP (what a coincidence) we are going to be left over. The manufacturer suggests that we use a 550W power supply to meet the needs of the system, as always we will check the actual consumption in our test bench.

In the side area we have the PCIe x16 connector in version 3.0. We must remember that the PCIe 4.0 bus has already been implemented for AMD X570 chipset boards and AMD Ryzen 3000 series CPUs. In any case, this type of bus will not be useful for current graphics cards. Finally, indicate that in this super version, we also do not have a Virtual Link connector for multi GPU, it does not take us by surprise anyway.

Nvidia GeForce RTX 2060 Super Hardware and Components

We leave the exterior design behind to tell you in detail which components are chosen for this new evolution of the 2060 with the heart of 2070. In this case we have not removed the heatsink to see its internal components, for the simple reason that we are going to find pretty much the same as on an RTX 2070.

After seeing the specifications that we give below, no one will be surprised that we are dealing with a modified version of the TU106 chipset that is used in RTX 2070 graphics cards. A chip with a 12nm Turing FinFET architecture that runs at a base frequency of 1470 MHz and 1650 MHz in turbo mode, being slightly higher than the frequency of the RTX 2070's TU106.

A total of 2176 CUDA cores have been implemented in this core, compared to 1920 for the RTX 2060, along with 272 Tensor cores and 34 RT cores. They will be in charge of processing the entire ray tracing system in real time (Ray Tracing) and DLSS (Deep Learning Super Sampling), the evolution of traditional Antialising. Nvidia still has the exclusivity of being the only one to have these two technologies on the market. And supporting this chip we have a L1 cache of 2, 176 KB and a L2 cache of 4, 096 KB, figures much higher than we had before. All this generates a capacity of 64 ROPs and 136 TMUs, at speeds of 6 Giga Rays / second, 57.4 TFLOPS, and 7 + 7 TOPS in FP32 and INT32 respectively.

And in the memory section we also have enough news, and once again coincide with the RTX 2070. The capacity has been increased to 8 GB GDDR6 working at 14 Gbps with a 7000 MHz memory clock. The bus has been increased to 256 bits to now deliver a speed of 448 GB / s, ranking alongside the RTX 2070 and 2070 Super. This great increase in benefits, will make performance far behind the normal 2060 version, which is what the manufacturer intends, directly competing and beating the AMD RX 5700 XT?

As for the cooling system, Nvidia series limits this GPU to work below 70 ° C, where the fans reach around 1770 revolutions per minute. For users who want to overclock, the TDP can be raised up to a maximum of 120%, and the RPM of its fans up to 3700 RPM. Obviously we would have a pretty noisy system.

Given its benefits, we can not strictly say that a copy-paste of the TU106 of 2070 has been made, although we already say that it is the same. And it is that we have a slightly lower count of processing cores and also of cache memory. This causes having a slightly lower TDP with the consequence of yielding a little less in pure IPC and presumably a reduced capacity to overclock the chip and its memory. The manufacturer may have chosen to disable the remaining cores and cache KBs until adjusting the performance to the desired, we do not complain either, it is a practice that is done in all current CPU chiplets.

The most positive thing about all this is that now we will have more graphics cards to choose from, and it is also assumed that it will be at prices similar to or lower than their "close" versions, so to speak, of normal RTX. What causes this? Well, that the normal RTX come down in price, we are talking about the RTX 2080 and perhaps the 2080 Ti, but with a disadvantage, and that is that now the RTX 2080 and RTx 2070 models will be discontinued. And let's not kid ourselves, it is a clever maneuver by Nvidia to put in check the two new GPUs released by AMD at the same time as these RTX, presumably exceeding in price and performance. We will see if the play works out well for them.

Test bench and performance test

Next, we are going to dedicate ourselves to perform the entire battery of performance tests on this Nvidia GeForce RTX 2060 Super.

TESTING BENCH
Processor:	Intel Core i9-9900K
Base plate:	MSI MEG Z390 ACE
Memory:	G.Skill Sniper X 16 GB @ 3600 MHz
Heatsink	Corsair H100i RGB Platinum SE
HDD	ADATA Ultimate SU750 SSD
Graphic card	Nvidia RTX 2060 Super
Power supply	Be Quiet! Dark Power Pro 11 1000W
Monitor	Viewsonic VX3211 4K mhd

All the synthetic tests and tests have been carried out with the filters to the maximum, as they come in the configuration of each software. The tests consist of tests that run in various resolutions, such as Full HD and 4K. We have run all of them on the Windows 10 Pro operating system in its 1903 version with the latest version drivers available for this graphics card (they have provided us with new ones before launching them for sale).

What are we looking for in the tests?

First, the best possible image quality. The most important value for us is the average FPS (Frames per second), the higher the number of FPS the more fluid the game will go. To slightly differentiate the quality, we leave you a table to assess the quality in FPS, but we will also have the minimum FPS in the tests that were thus possible:

FRAMES PER SECOND
Frames Per Second (FPS)	Gameplay
Less than 30 FPS	Limited
30 ~ 40 FPS	Playable
40 ~ 60 FPS	Good
Greater than 60 FPS	Fairly Good or Excellent

Benchmarks and synthetic tests

For benchmark tests we will use the following titles:

• 3DMark Fire Strike normal3DMark Fire Strike UltraTime SpyPort Royal (RT) VRMARK

Game testing

After the synthetic tests, we will proceed to evaluate the real performance in games, thus having a closer guide of what our GPU will be able to deliver under DirecX 12, 11 and OPEN GL.

The tests will be carried out at the three most used resolutions in gaming, we refer to Full HD (1920 x 1080p), QHD or 2K (2560 x 1440p) and UHD or 4K (3840 x 2160p). In this way, we will have a complete range of results to be able to compare them with other GPUs. For each of the games, we have kept the automatic settings selected in each and for each resolution. And to expand the information, we are going to leave the performance records with both a normal configuration and another with Ray Tracking + DLSS in the games that are possible.

The games tested, along with their quality, texture filter and execution API.

• Final Fantasy XV, standard, TAA, DirectX 12DOOM, Ultra, TAA, Open GL 4.5 Deus EX Mankind Divided, Alto, Anisotropico x4, DirectX 11Far Cry 5, Alto, TAA, DirectX 12Metro Exodus, Alto, Anisotropico x16, DirectX 12 (with and without RT) Shadow of the Tomb Rider, Alto, TAA + Anisotropic x4, DirectX 12 (with and without DLSS)

After seeing the results, we have a considerable performance increase compared to the normal RTX 2060, in fact we have results that exceed 20 FPS in some games. The trend in the results against the Nvidia RTX 2070, is to match them and even exceed their records in many of the games. Despite being a chipset cut in performance to that of the normal 2070, the truth is that the optimization of drivers and in games make it equal and exceed this one.

Gaming performance with DLSS + Ray Tracing enabled

We have carried out the same test for Shadow of the Tomb Rider games and especially for Metro Exodus, which allow us to use the new technology of the Nvidia RTX to compare the FPS obtained. The rest of the graphic options have been kept exactly the same.

	1920 x 1080 (Full HD)	2560 x 1440 (WQHD)	3840 x 2160 (4K)
Metro Exodus (without RTX)	75 FPS	58 FPS	34 FPS
Exodus meter (with RT + DLSS)	65 FPS	49 FPS	23 FPS
Shadow of the Tomb Rider (without RTX)	119 FPS	86 FPS	49 FPS
Shadow of the Tomb Rider (with DLSS)	114 FPS	91 FPS	59 FPS

Overclocking

As always, we have overclocked this Nvidia GeForce RTX 2060 Super to see where it is capable of going with a rise in laps using the MSI Afterburner software .

Metro Exodus (without RTX)	Stock	@ Overclock
1920 x 1080 (Full HD)	75 FPS	81 FPS
2560 x 1440 (WQHD)	58 FPS	63 FPS
3840 x 2160 (4K)	34 FPS	38 FPS

After having been testing the stability of the GPU making modifications to the core clock and memory, in addition to slightly increasing the power of the ventilation system, we have managed to squeeze this card a little more. In our case, reaching a maximum of 1800 MHz in the processor and 7200 MHz in GDDR6 memory.

It's not a sizeable rise in terms of pure clock cycles, but the idea is to strike a balance of stability in the GPU so that the boost is reflected in a performance boost as well. We have not touched the voltage of the GPU, but we have touched the cooling profile of the fans to ensure a good temperature. With the standard profile, the GPU limits its performance to 70 ° C with the fans at 1770 RPM, while removing this limit, we can go up to 88 ° C and no less than 3700 RPM of its fans.

Temperatures and consumption

In addition to measuring both its temperature with the HWiNFO program by stressing the GPU with FurMark, we have also simultaneously measured the power consumption of the entire equipment. And while we were doing it, we have taken a few thermal captures with the card at full capacity with an ambient temperature of around 24 ° C.

Here we can see that the Nvidia GeForce RTX 2060 Super in its stock configuration will not exceed the neck of 70 ° C, which is just what we discussed during overclocking. Similarly, consumption falls within expectations, always remembering that it is from the complete equipment (except the monitor). If we also added Prime95 to the CPU, we would get about 307W of consumption.

Final words and conclusion about Nvidia GeForce RTX 2060 Super

Well, we come to the end of this review of the new Nvidia GeForce RTX 2060 Super, and we have seen results that are in line with what was promised, in fact, they even exceed expectations a little by being better than the previous RTX 2070. The inclusion of the TU106 chip, although a little cut in cores, does not diminish the extraordinary power of this Super version.

Precisely this performance has left us frame rates per second greater than 100 in Full HD resolution and greater than 60 FPS in 2K, even very close to 50 FPS in 4K resolution. So we can practically play everything in high and ultra quality with practically no fluency problems. Similarly, playing with RT and DLSS activated will not be a problem in these first two resolutions.

As for the overclocking capacity, we can determine that it is good, although not excellent. We have been able to turn the memory and clock of the GPU up to around 1800 MHz and 7200 MHz respectively. Something enough to boost up to 7 FPS Metro Exodus in Full HD, not bad.

We recommend our guide to the best graphics cards on the market

And the cooling system continues to work like a charm, with a double system of 85mm fans that are capable of reaching 3700 rpm if we overclock. In stock configuration it is a fairly silent system, although the solution of turning off the system when the card is idle has not been implemented. And this is what we consider of great importance today.

It is undoubtedly a card that can practically go through its one RTX 2070, although with minor adjustments to match its price and minor aesthetic improvements. Clearly built to compete with the AMD RX 5700 XT that will also appear on the same day on the market, although of course adding DLSS and RT, something that AMD does not yet have.

This reference model Nvidia RTX 2060 Super will be available at its output for a price of 419 euros, being somewhat expensive than the price at which the RTX 2060 came out in its day. Considering that the RTX 2070 Super will appear at a price similar to that of the RTX 2070, in this case they should have done the same. From all appearances, 419 euros for a card equal to or greater than 2070 is still interesting, but a cost equal to the RTX 2060 would have been correct. In addition, the normal RTX 2070 model will be discontinued.

ADVANTAGE	DISADVANTAGES
+ EXCEEDS OR EQUALS PERFORMANCE TO RTX 2070	- IT IS STILL A MODIFIED RTX 2070
+ EXCELLENT DESIGN, AS ALWAYS	- THE FANS CONTINUE WITHOUT STOPPING WITH THE SYSTEM AT REST
+ OVERCLOCKING WITH VERY GOOD RESULTS
+ IDEAL FOR PLAYING IN FULL HD AND 2K WITH AND WITHOUT RT
+ EFFICIENT AND EXCELLENT TEMPERATURES

The professional review team awards him the gold medal

Nvidia RTX 2060 Super

COMPONENT QUALITY - 90%

DISSIPATION - 90%

GAMING EXPERIENCE - 87%

SOUND - 91%

PRICE - 84%

88%

The performance of the RTX 2070, but the price is high