We Tested and Selected the Best Graphics cards for Battlefield 5 (V) in terms of Performance, Price, Heating / Noise levels and more.
Above you can see the Ranking of the Best Graphics cards for Battlefield V and below you will find additional info about the Game and in-depth reports of each Graphics card.
Guide: Battlefield 5 (V)
Battlefield 5 is a return to the roots of the series, namely the setting of the Second World War. And if you played the last two Battlefield parts 4 and 1 (we can’t do anything for the confusing numbering!), you can already guess what to expect from Dice’s new shooter: Multiplayer wow, singleplayer poo.
Because while the Swedish developers always unpacked the really big game fun trowel in the multiplayer modes of the two predecessors, the solo campaigns turned out to be maximally solid appendages despite good approaches, which were entertaining, but in no way memorable or even playfully outstanding.
And to make a long story short: Battlefield 5 fits seamlessly into this group. So we’d advise anyone who wants to buy the title just for the solo mode to be careful.
The concept of the single-player campaign is similar to that of Battlefield 1, so the missions don’t tell a big coherent story, but three independent “war stories”, all with their own little storyline and main characters.
In “Under No Flag” you are sent into the North African desert as a militarily inexperienced ex-bank robber to sabotage German positions.
“Nordlys” puts you in the role of a young resistance fighter who uncovers the plans of a German weapon of retaliation, and in “Tiralleur” you fight as a member of a French colonial troop against the German occupiers in France.
In addition, there is an only partially playable intro mission, which, however, is finished after barely 15 minutes.
Ranking First: NVIDIA GEFORCE RTX 2080 Ti Founders Edition
- Great performance advantage over the GeForce GTX 1080 Ti
- 4K / VR easily possible
- Low temperatures and noise
- High-quality current and voltage supply
- Expensive, but the best performance always costs
How powerful the two models GeForce RTX 2080 Ti and GeForce RTX 2080 are, where the strengths and possibly the weaknesses lie, we’ll now look at.
Already last week we reported in a detailed article about the Turing architecture and the new RT and Tensor Cores. With the Turing architecture, NVIDIA has made numerous changes to the design of the streaming multiprocessor, combining this with faster GDDR6 memory.
Innovations like Mesh Shading and Variable Rate Shading are supposed to speed up the well-known rasterization process and before you start working with the new Hybrid Rendering Pipeline of GeForce RTX cards, you should understand the working principle of the different ray tracing techniques.
A GeForce RTX graphics card for Battlefield 5 (V) without considering the RTX, tensor and RT cores as well as the other innovations is not possible or does not do justice to the matter.
In the context of this article, we will also discuss Deep Learning Super Sampling (DLSS) and the NV Scanner as a new possibility of overclocking, besides the pure performance data.
All of them are of course based on the new Turing GPUs. Their production takes place in 12 nm at TSMC. With an area of up to 754 mm² and up to 18.6 billion transistors, the chips are among the largest ever produced and are also by far the largest in their performance class that have ever been used on an end-customer graphics card for Battlefield 5 (V). We also conducted a more detailed analysis of the production of the GPUs last week.
The GeForce RTX 2080 Ti offers 4,352 shader units, the GeForce RTX 2080 and GeForce RTX 2070 come up with 2,944 and 2,302 respectively on correspondingly less.
One RT core per SM cluster means a number of 68, 46 and 36 for the three models, and there are eight tensor cores per SM cluster, so that we arrive at 544, 368 and 288 here.
The number of geometry and texture units also depends on the respective expansion stage. The individual clock rates will still play a role when we take a closer look at the two maps. With the Turing GPUs and the new cooling the behaviour of the GPU boost mechanism changes.
The switch to the faster GDDR6 memory with an equally large memory interface of 352 or 256 bits pays off for NVIDIA in the form of higher memory bandwidth.
The GeForce RTX 2080 Ti Founders Edition is NVIDIA’s flagship product, but poaches where the Titan cards were previously positioned. Accordingly, we of course compare the new model with the GeForce GTX 1080 Ti Founders Edition, but always have to keep the new positioning in mind.
In addition to the new hardware functions of the Turing architecture, the GeForce RTX 2080 Ti also offers 4,352 shader units and thus 21% more than its predecessor, the GeForce GTX 1080 Ti. The memory bandwidth grows by 27% to 616 GB/s due to the use of GDDR6 memory. The memory expansion remains identical with 11 GB.
The GPU’s clock rates aren’t that much higher on paper, the question is whether the GeForce RTX 2080 Ti Founders Edition is able to maintain this boost clock rate due to the new cooler or whether we’ll often see a throttling, as in the GeForce GTX 1080 Ti Founders Edition, due to a temperature limit and an undersized cooling.
We’ll go into the details of the power supply of the new cards a bit more in detail. But now we want to take a closer look at the new models.
Even though NVIDIA has decided on a completely new cooler design, the basic features of a graphics card for Battlefield 5 (V) are of course retained.
With a length of 268 mm, NVIDIA stays within the ATX specifications and the height of two slots is no problem even in more compact cases. The two fans have a diameter of 85 mm and do not switch off in idle mode – but more about that later.
For the first time in years, NVIDIA is changing something about the basic principle of cooling their own cards. Instead of a radial fan that blows air through an enclosed cooler housing, the new models use a dual axial fan design.
This means that two fans push the air from the top towards the PCB and thus onto the cooler. What NVIDIA’s partners have been doing for years is now being applied to the Founders Editions.
In addition to the new arrangement and use of two fans, the appearance also changes slightly. A lot of aluminum has been left, but now a black cover between the two fans is added to the front. The new cards are once again of higher quality than the previous models and also significantly heavier.
There is a black cover between the two axial fans, and since there is no more room for the model name of the card underneath the slot bracket, NVIDIA has placed it between the two fans.
The two axial fans have a diameter of 85 mm each and have 13 fan blades. These almost completely close off the fan’s opening and thus do not correspond to a rather open design with narrow fan blades.
NVIDIA still doesn’t plan for the fans to stand still in idle mode, but they want to be much quieter in idle mode as well as under load than was the case with the previous Founders Editions.
The fan speeds range from 1,500 rpm in idle mode to 2,000 rpm under load.
The obligatory GeForce logo is used on the front of the card, but now has the addition of RTX instead of GTX. An LED also illuminates the logo. It seems that NVIDIA has installed RGB lighting here, which could theoretically also be controlled in other colors. So far, however, the lettering shines in the familiar green.
The two axial fans are surrounded by a projecting frame. This emphasizes the appearance of the two existing “shafts” for the fans. On the sides there are some more aluminium panels, whose edges have been beveled once more, which makes them shine. You can certainly blame NVIDIA for a lot in terms of pricing, but the cooler makes a very high-end impression and can compete with everything we’ve seen from the partners so far.
The additional power supply for the card is provided by two 8-pin connectors. These supply 150 W each according to the PCI Express specification.
Together with the 75 W of the PCI Express slot, we thus arrive at a supply of 375 W. NVIDIA provides for a thermal design power of 260 W, but gives itself a bit of room to overclock itself to allow the user to overclock.
We’ll go into the details of the power and voltage supply once again in more detail, as the changes here are also far-reaching.
A dual-axial design for cooling naturally changes the principle of airflow. The air is no longer blown in the direction of the slot bracket and out of the housing, but remains there as far as possible.
The two fans press the air onto the heatsink and the heatsink guides the air up and down and out of the side of the cooler. For this reason NVIDIA has left more or less large openings around the card.
The back of the card is equipped with a bright aluminium backplate. Due to the radial fan, this could be partially removed when two or more cards were used in a system. This is no longer possible with the GeForce RTX 2080 Ti Founders Edition. The backplane ensures mechanical stability of the card and protects it from damage.
NVIDIA also strives for a high-quality look on the back side and decorates the card with the model name. This is encased by longitudinal grooves in the backplate.
Unfortunately, we can’t yet come up with SLI tests at the start of the GeForce RTX 2080 Ti Founders Edition. The card has an NVLink connector on the front side of the card.
This is used for communication of an SLI system. A cover protects it and when it’s not in use, the cover provides a uniform look.
The slot bracket of the GeForce RTX 2080 Ti Founders Edition only has a few openings left. Since the fan doesn’t blow the air directionally anymore, there are probably no more openings necessary.
Display outputs are HDMI 2.0b once and DisplayPort 1.4 three times. NVIDIA is also the first graphics card manufacturer to install a VirtualLink in the form of a type C connector. More details about the new video and display engine and VisualLink can be found in the following.
NVIDIA has completely sealed the other end of the GeForce RTX 2080 Ti Founders Edition. The two axial fans are supposed to blow out the warm exhaust air to a large extent on the front side of the card, since ventilation inside the case can also be ensured here.
On the following page we have separated the GeForce RTX 2080 Ti Founders Edition from its cooler and take a look at it and the PCB of the card.
At first glance, the PCB of the GeForce RTX 2080 Ti Founders Edition does not offer any special features. Since no High Bandwidth Memory is used, we already know the rough structure of the PCB.
In the middle you can see the GPU. The 11 GDDR6 memory chips are placed as equally spaced as possible. To the left and right of them the power supply of GPU and memory can be seen.
The TU102-GPU used in the GeForce RTX 2080 Ti Founders Edition comes on a chip area of 754 mm² and this is very well visible on the picture also in comparison to the other components.
Also recognizable are the tracks that run from the GPU package to the individual chips. These are connected to the GPU with 32 bits each.
With a 352 bit wide memory interface and 32 bits per chip we get 11 GDDR6 chips. So one space remains free and is used for example with the Quadro RTX 6000 and Quadro RTX 8000.
The GDDR6 memory chips on the GeForce RTX cards are currently all from Micron and are named MT61K256M32JE-14:A. These offer 8 GBit/1 GB per chip and operate at a voltage of 1.35 V.
The rear third of the PCB belongs to the controlling and management of the current and voltage supply. Here we see for example the two 8-pin connectors. On the left you can already see some of the 13+3 voltage phases used in the GeForce RTX 2080 Ti Founders Edition.
Ten of the total 16 voltage phases are on the right side of the GPU and here in the middle of the picture. NVIDIA does without phase doubling and addresses the 16 phases individually.
According to NVIDIA, phase doubling has too many disadvantages and also makes the design less flexible, which is why they decided on the more complex solution, which also saves space on the PCB.
A 13+3 phase design describes the supply of the GPU by means of 13 phases and the other three take care of the memory. To increase efficiency, NVIDIA can switch the voltage phases on or off depending on the load. Why this is done is explained on a separate page about power and voltage supply.
The GPU built into the GeForce RTX 2080 Ti Founders Edition is named TU102-300A-K1-A1. The TU102-300 describes the expansion stage. The ending “A” indicates according to previous reports that a GPU is used here, which was selected.
Models of the GeForce RTX 2080 Ti without overclocking should use the GPU with the designation TU102-300-K1-A1 and thus without “A”.
Once again a look at the rear end of the PCB. NVIDIA did not only control the 13+3 voltage phase of the GeForce RTX 2080 Ti Founders Edition, but also the balancing of the supply via the additional two power connectors.
This should also result in the new Turing cards being significantly more efficient than their predecessors and also allow for a slightly higher overclocking.
We have already mentioned the highly visible traces to and from the eleven GDDR6 memory chips. On the picture above you can see the tracks of the PCI Express interface, which are led around the GDDR6 chips.
The view to the back of the PCB is much less spectacular. Very nice are the backside areas of the GPU and the voltage phases. NVIDIA covers the backside with a very thin aluminum backplate, which should fulfill its purpose as protection.
While the front side of the PCB is also populated with larger components, the back side contains almost only small SMD components. However, these are used thousands of times. The picture certainly makes the complexity of the PCB of a modern graphics card for Battlefield 5 (V) particularly clear.
The GeForce RTX 2080 Ti Founders Edition is cooled by a large front panel on which all components except the GPU are mounted.
Thermal pads are supposed to improve the contact between the components and the frontplate. The actual heat sink is located behind this frontplate. It uses numerous heat pipes to better distribute the waste heat.
The GPU itself sits on a base plate, which in turn is directly connected to a vapor-chamber. According to NVIDIA, the GeForce RTX 2080 Ti Founders Edition uses the largest vapor-chamber ever used on a graphics card for Battlefield 5 (V).
Instead of multiple 4-pin connectors, NVIDIA uses a ribbon cable on the GeForce RTX 2080 Ti Founders Edition to power the two fans and the LEDs and read a fan signal. Almost two dozen screws are needed to remove the cooler.
A disassembly of the cooler itself takes it to 70 screws and four of them alone are responsible for routing the ribbon cable.
Turing-OC and NV Scanner
The improved current and voltage supply should of course have positive effects on overclocking. The manual overclocking remains, but the voltage still limits NVIDIA upwards, as too high a voltage is the main reason for damage to a GPU.
However, manual overclocking, for example using EVGA’s Precision X1, has a few limitations. The voltage slider is no longer linked to a fixed value. Instead, the available voltage range is mapped as a percentage.
GPU-Boost 4.0 brings some changes in the clock behavior depending on the temperature. Graphics cards with Pascal-GPU reduced the clock speed very quickly when the temperature limit was reached. GPU-Boost 4.0 and the Turing GPUs on the other hand reduce the clock speed to the boost limit and then reduce it further if necessary.
But the manual overclocking of both cards will be discussed separately. Here we will talk about the new NV scanner, which should simplify the overclocking.
Together with the new cards of the GeForce RTX-20 series, NVIDIA introduces a new component of the NV API, which is used by programs like EVGA Precision and MSI Afterburner to allow control and overclocking of the cards.
The NV scanner is designed primarily to reduce the time needed for overclocking. The NV scanner is quite similar to the OC scanner of the Pascal generation.
Already for Pascal, NVIDIA built a method into the API to GPU Boost, where a clock/voltage curve was automatically run and the highest clock for a certain voltage level was determined.
However, the new NV Scanner API has its own load test, which should make overclocking much easier. Instead of testing the stability with one or more games, which can lead to the overclocking running stable with game A but not with game B, a so-called NV workload is used.
This workload consists of mathematical calculations and no 3D representation. It runs as a closed system and does not cause the complete computer or even the driver to crash. This of course shortens the time needed to test a certain OC value.
The NV Scanner uses five voltage points, runs the clock of the GPU on them and interpolates between these points the other values of the voltage/frequency curve.
For the GeForce RTX 2080 Founders Edition, the NV Scanner determined a GPU offset of +112 MHz to just over 2,000 MHz GPU clock. For the GeForce RTX 2080 Ti Founders Edition, the NV Scanner also determined almost 2,000 MHz GPU clock at an offset of +186 MHz.
The NV Scanner also worked with cards based on the Volta and Pascal architecture in our tests. We have the current versions of the EVGA Precision X1 as well as the MSI Afterburner and we were able to test this with both.
Regarding all RTX technologies, NVIDIA must now deliver. DLSS only works in two demos today, but no relevant game. At least we have the possibility to watch a ray tracing demo. This is the Star Wars Elevator demo, which already uses some ray tracing effects. In the version tested by us DLSS is also used.
We ran the demo on a NVIDIA Titan V and on the two new cards. However, there was a problem with the GeForce RTX 2080 in 2160p, so we don’t have any results here.
The Star Wars “Reflections” demo shows two Stormtroopers and Captain Phasma entering an elevator. Especially impressive are the reflections on the armor Captain Phasma.
In order to be able to calculate in real time, a workstation like the NVIDIAs DGX-1 with four GPUs was necessary in spring. But the goal is to execute such effects on a single graphics card for Battlefield 5 (V). An NVIDIA Titan V with Volta architecture achieved 9 or almost 13 frames per second.
We could only test the GeForce RTX 2080 Founders Edition in 1440p and reached 34.6 FPS here. A GeForce RTX 2080 Ti Founders Edition achieved 55 FPS at 1,440p, and almost 44 FPS at 2,160p.
Noise and Heating levels
With the new dual axial design, NVIDIA wants to make a big leap forward in terms of volume under load. In fact, we can attest the Founders Editions of GeForce RTX 2080 Ti and GeForce RTX 2080 a significantly lower volume than their predecessors.
With 40.6 and 41.1 dB(A), respectively, we are clearly below what we had to put up with the previous Founders Editions. However, some custom models will certainly be a bit quieter – at least this is what the value for a GeForce GTX 1080 Ti in custom design suggests. NVIDIA doesn’t plan to turn off the fans in idle mode.
We measured the power consumption in two ways. Once in the entire system and once only the cards themselves. The overall system therefore had a not to be sneezed at in the first measurements and the results look accordingly.
The values can be judged a bit better if only the card’s consumption is measured. Here, the GeForce RTX 2080 Ti Founders Edition reaches just over 260 W. Thus, it consumes 15 W more than the GeForce GTX 1080 Ti in the Founders Edition. The top model is still more economical than the fastest Vega variant, though.
GeForce RTX 2080 Founders Edition has an increased consumption compared to its nominal predecessor. However, we are talking about a plus of 55 W here. Whether and to what extent the new cards are more efficient can only be estimated when we also know the performance.
The temperature measurements show that both cards operate far from the temperature limit. The GeForce RTX 2080 Founders Edition doesn’t get warmer than 73 °C and the GeForce RTX 2080 Ti Founders Edition only reaches 77 °C. This means that NVIDIA can keep another promise regarding cooling in addition to volume.
A few words about the coil whistling: We couldn’t detect any. NVIDIA has, according to their own statement, also tried to do without coils as far as possible in the power supply.
Of course, this is only possible if the respective circuit can be implemented without coils. Whether the missing coil beeps can be transferred to the other cards remains to be seen.
Turing, Volta and Pascal at 1,850 MHz
To estimate the performance of the Turing architecture in comparison to the Pascal and Volta architecture, we ran the two GeForce RTX cards as well as a Titan V and two GeForce GTX cards with a GPU clock of 1,850 MHz each and sent them through the benchmarks.
It is noticeable that an NVIDIA Titan V can keep up more than well. In some cases, it is on par with the GeForce RTX 2080 Ti Founders Edition – or even outperforms it.
The reason is obvious. The GeForce RTX 2080 comes with 4,352 shader units, a Titan V has 5,120. The memory bandwidth is almost identical with 616 to 652.8 GB/s and shouldn’t have a big impact. However, almost 18% more shader units at the same clock rate do have their influence and make sure that the Titan V shows an impressive performance.
Besides the automatic overclocking by means of NV scanner, we have of course also tried manual overclocking. For this we used the EVGA Precision X1 tool and played a little with the knobs.
The power limit can be set to 123% in the GeForce RTX 2080 Ti Founders Edition. This means an increase from 260 to 320 W.
The standard voltage is read out at 0.994 V and can be increased to 1.050 V. So there is not much leeway here. We used the power and voltage limit to the maximum and ended up with a stable clock rate of 2.125 MHz. We could overclock the memory from 1.750 to 1.900 MHz.
We can also increase the power limit to a maximum of 123% for the GeForce RTX 2080 Founders Edition. Here this means an increase of the limit from 225 to 278 W. For the manual overclocking, however, the end of the line was already reached at 2,100 MHz GPU clock.
We had expected a bit more in comparison to the larger model. We were able to increase the memory clock from 1,750 to 1,900 MHz again. We turned up the fans a little bit each time so that the temperatures remained in about the same range as before.
Especially the plus in the GeForce RTX 2080 Ti Founders Edition is more than remarkable. However, the power consumption also increases rapidly with the overclocking. The GeForce RTX 2080 Founders Edition shows a more moderate overclocking potential, but still the free performance plus is also remarkable here.
Verdict and Performance
We have to go a bit further to sum up the GeForce RTX 2080 Ti and GeForce RTX 2080 Founders Edition. With the GeForce RTX 20 series, NVIDIA goes beyond what a “simple” rasterizer GPU wants to be. But objectively speaking, the performance in this area must also fit. Focusing only on RTX, ray tracing and more will not convince most buyers.
And exactly in this direction we have tried to develop the theory of Turing architecture. NVIDIA has reworked the shaders, which, like the Volta architecture, means that we see a higher performance advantage in low-level APIs than in classic DirectX 11 applications.
This is also due to the fact that NVIDIA introduces new features such as mesh shading or variable rate shading, which can be implemented quite easily by developers and in some cases brings a significant performance advantage over previous rendering technologies.
But before we come to the evaluation of the two cards, a few words about RTX, DRX, Ray Tracing, NGX and DLSS: NVIDIA is ultimately one of many companies that wants to sell new products.
However, in the GPU computing sector, the past few years have already shown that the right idea at the right time can be worth a lot – NVIDIA is growing particularly strongly in the data center segment.
The step towards hardware-accelerated ray tracing and the implementation of deep learning applications in this segment is to be seen as a positive development. While the NGX framework is a pure implementation of NVIDIA, ray tracing using Microsoft’s DirectX Ray Tracing (DXR) can be a real innovation for gamers, which is reflected in a broader implementation than we’ve seen with many other effects.
NVIDIA uses the RT Cores to hardware accelerate Microsoft’s DXR interface. This is open to all hardware manufacturers and AMD and Intel can also address it. You should also keep this in mind.
The GeForce RTX 2080 Ti Founders Edition will be and remain the top model in the coming months. A “Titan Xu” or whatever is denkar because of the larger expansion of the TU102 GPU, but this time it is rather unlikely.
This is also due to the fact that NVIDIA already charges 1,250 Euro for the GeForce RTX 2080 Ti Founders Edition, although a Titan V also cost 3,100 Euro.
All in all, the performance plus of the GeForce RTX 2080 Ti Founders Edition is a bit less in comparison to the GeForce GTX 1080 Ti Founders Edition than the smaller models. We’re also talking about up to 50%, but in return the deflections downwards are somewhat larger and depending on the game, as for Battlefield 5 (V) and resolution we see an increase of 17%.
In comparison to the GeForce GTX 980 Ti, the increase is again a factor of two with fluctuations up and down. The new top model is about 20% faster than the GeForce RTX 2080 Founders Edition. This naturally plays an important role in the price difference of 400 euro (1,250 to 850 euro).
The power consumption of the GeForce RTX 2080 Ti Founders Edition is only slightly higher, also because a thermal design power of 250 W is only reluctantly exceeded.
Here, a limit has been reached to a certain extent and all gains in performance can be attributed to efficiency improvements. The same applies to volume and temperatures for the GeForce RTX 2080 Ti Founders Edition as we have already noted for the smaller model.
The temperature is no longer the limiting factor; instead, the two cards now work at the power limit, but easily call up a boost clock rate that is far above the minimum specifications. The volume is only slightly higher and can be judged much better than with the previous FE designs.
All in all the GeForce RTX 2080 Ti Founders Edition is one of the best GPUs you can get right now, which is why it is ranking First on the Best Graphics cards for Battlefield 5 / V.
Ranking Second: Gigabyte GeForce RTX 2070 Super Gaming OC
- Good performance with Gaming
- Best Price
- 1.900 MHz OC Clock rate
- A bit noisey while heavy Gaming
The first overclocked RTX 2070 in test: Gigabytes RTX 2070 Gaming OC is equipped with Nvidia’s faster TU106-A graphics chip and has to prove itself against the unoverclocked MSI RTX 2070 Armor 8G.
With the Geforce RTX 2070 Gaming OC 8G from Gigabyte in the test, the first model of an RTX 2070 with the more powerful TU106-400A chip has arrived in the editorial office.
The 2070 custom design has to prove itself against the MSI RTX 2070 Armor 8G with non-A chip and reference clock rates, which we tested at the start of the third and so far cheapest Turing graphics card for Battlefield 5 (V) from Nvidia.
Due to the built-in TU106-400-A graphics chip on the RTX 2070 Gaming OC, Gigabyte can choose higher clock rates ex factory than MSI can on the RTX 2070 Armor 8G. The base clock remains identical as usual at 1.410 MHz, but the boost clock increases from 1.620 MHz (Nvidia reference clock) to 1.725 MHz.
With the help of Gigabytes Aorus Engine tool, the OC mode can additionally be activated, which raises the clock rate to 1.740 MHz. Then the Gaming OC clocks 30 MHz higher than Nvidia’s Founders Edition, which also relies on the more clock-friendly A version of the graphic chip.
While the power connections (1x 8-pin, 1x 6-pin) are the same with the MSI RTX 2070 Armor, the non-A GPU has to manage with 175 watts TDP, Gigabyte can increase the power limit to 185 watts thanks to the faster chip.
Other manufacturers such as Zotac with the RTX 2070 AMP Extreme even increase the TDP to 225 watts and can increase the clock rates even further.
Identical to the MSI RTX 2070 Armor, the RTX 2070 Gaming OC also has 8 GByte GDDR6 video memory with an effective clock rate of 14 GHz.
The RTX 2070 Gaming OC uses Windforce 3X cooling in a two-slot high design. All four copper heatpipes have direct contact with the graphics chip and transport the waste heat to the fine fins of the aluminium radiator.
Fresh air is provided by three 80 millimetre axial fans, which are stationary in Windows mode. In order to optimize the airflow and counteract turbulences, Gigabyte relies on so-called “alternate spinning”, whereby the outer fans run counterclockwise while the inner fan rotates clockwise.
The RTX 2070 Gaming OC’s most striking feature is probably its appearance. Gigabyte gives both the plastic radiator cover and the metal backplate a coat of bright white. In combination with the matt black fans, the model has a rich contrast ratio. However, the haptics suffers somewhat due to the choice of materials.
In our gaming benchmarks, the more powerful TU106-400-A graphics chip of the Gigabyte Geforce RTX 2070 Gaming OC in UHD resolution clocks with an average of 1,880 MHz. Compared to the MSI Armor and its 1,725 MHz, the custom design of Gigabyte clocks 155 MHz higher.
The supposedly low clock rate increase is surprisingly noticeable, depending on the game and resolution, with an additional performance of three to nine percent in Battlefield 5 (V), at least in the benchmarks. The difference is especially big in The Witcher 3: Blood and Wine: In WQHD resolution the frame rate increases from 90.7 to 96.4 fps.
On average, the overclocked Gigabyte RTX 2070 Gaming OC is about five to six percent faster than the MSI RTX 2070 Armor, which gets by with standard clock rates.
Due to similar clock rates and dimensions, we suspect that Nvidia’s Geforce RTX 2070 Founders Edition should be as fast as Gigabyte’s custom design, but Nvidia doesn’t release samples of the 2070 FE and leaves the field to the custom designs.
With these test results, Gigabytes 2070 Custom Design moves significantly closer to the GTX 1080 Ti.
However, this only counts for the GTX 1080 Ti Founders Edition, which acts about ten percent slower than overclocked and with massive coolers equipped custom designs of the 1080 Ti due to reference clock rates and the traditional cooling design including radial fan.
Heating and noise levels
The Gigabyte RTX 2070 Gaming OC can also distinguish itself positively for the most part in the other test disciplines. The custom design is silent in Windows mode, the chip heats up to 42 degrees.
Under gaming load, the three fans rotate with around 2,000 revolutions per minute (62 percent fan power) and produce an audible 40.8 decibels. A still good value, especially in view of the only two slot high cooling design and the chip temperature of 71 degrees.
In comparison to the three slot high MSI RTX 2070 Armor, which is equipped with two 100 millimeter fans, the significantly more compact Gaming OC is, as expected, the shortest in terms of volume.
The measured values for volume seem to differ only slightly at first glance, but only the MSI remains really quiet even under longer load. And MSI also positions itself clearly in front of Gigabytes Gaming OC with a load value of only 64 degrees in terms of temperatures.
In terms of power consumption, the RTX 2070 Gaming OC proves to be a bit hungrier due to the higher TDP. While we measured identical rates of 43 watts in idle, the power consumption for the entire test system increases to 294 watts under load.
Thus, the Gigabyte Gaming OC requires four percent more power with a roughly six percent higher performance compared to MSI’s RTX 2070 Armor.
The Gigabyte Geforce RTX 2070 Gaming OC illustrates Nvidia’s two-class society in the test: On the one hand, the manufacturers have the choice to buy less clocked graphics chips at a lower price and offer corresponding models with a TDP of 175 watts starting at 519 euro.
Or they decide to install the more overclockable chips (TU106-400A) on their graphics cards for Battlefield 5 (V), increase the clock rate and TDP ex works and charge a slightly higher price for it.
Gigabyte’s custom design has such an A-GPU and can set itself apart from the MSI RTX 2070 Armor 8G by an average of six percent in Battlefield 5 (V)due to a high chip clock of just under 1,900 MHz.
As expected, the performance increase turns out measurable, but not noticeable in gaming practice, and the Gaming OC should perform similar to Nvidia’s Founders Edition due to its dimensions and clock rates.
Higher prices are demanded not anymore for RTX 2070 models, such as the Gigabyte Gaming OC, due to the faster graphics chips for Battlefield 5 (V), but currently both model variants are quite close to each other in terms of price, so that it’s worth reaching for higher clocked custom designs.
Especially since the models with standard chip delivered with reference clock rates have little room for higher clock rates, as the power limit limits very quickly.
All in all the Gigabyte RTX 2070 Gaming OC is ranking second on the Best Graphics cards for Battlefield 5 / V. Due to its relatively low price, it also won our ‘Best Price’ category, so if you are on a budget, this is the graphics card for you!
Ranking Third: Gigabyte GeForce RTX 2060 Super Gaming OC
- Great performance with Gaming
- Good Price in 2020
- Very Silent
- Cooling okay but not great
The GeForce RTX 2060 Super is Nvidia’s entry into the new “Super” model series. The graphics card for Battlefield 5 (V) rounds out the original GeForce RTX 2060 with around 15 percent more performance and an increased memory of 8 GB.
Nvidia’s own Founders Edition is doing quite well, but still offers potential for improvement, especially in the cooler. The test of two custom models should clarify whether they can do it better.
Gigabyte has three different versions of the GeForce RTX 2060 Super in its portfolio. The tested GeForce RTX 2060 Super Gaming OC is the middle model, which changes hands from 453 euro. Despite the name affinity to the GeForce RTX 2070 Super Gaming OC from the same manufacturer, there is a bigger difference.
Thus, not the 2.5-slot Windforce cooler is used, but the weaker 2.0-slot variant. This saves space, but there is also less cooling material. The cooler itself was also purged.
The base still consists of an aluminum core with various fins, but instead of six heatpipes, the TU106-GPU only has four instead of six. The three axial fans were also reduced in size. The diameter falls from 80 to 75 mm and the fans are also flatter.
The fan behavior is the same, though, because the fans switch off at GPU temperatures of less than 62 degrees Celsius and the graphics card works silently.
An eight-pin and a six-pin power connector must be connected to operate the GeForce RTX 2060 Super Gaming OC. The graphics card may require 215 watts, but the power consumption can be manually increased up to 280 watts – strangely enough, this is 40 watts above the actual maximum of the GeForce RTX 2070 Super Gaming OC.
The base clock of the graphics card for Battlefield 5 (V) is 1.470 MHz, the average turbo is specified with 1.815 MHz. That’s 165 MHz more than Nvidia plans for the Founders Edition, but its specification is chosen very low. The 8 GB GDDR6 memory works with the normal 7.000 MHz.
The graphics card offers RGB, but this is more a checklist feature than a useful lighting. Because there is nothing more than the gigabyte lettering on the page, which also shines rather weakly and can’t display some colors at all.
Both graphics cards clock with a maximum of 1,935 MHz under full load, as long as neither the power consumption nor the temperature limit. However, both products are configured so that they almost always run into the power limit and then clock down.
For example, you don’t see the full 1,935 MHz on the Gigabyte GeForce RTX 2060 Super Gaming OC at any time. There isn’t more than 1.905 MHz, which are then at least occasionally present in most games.
Metro Exodus is the worst case for the graphics card. In the first-person shooter the GPU works alternately with 1.860 MHz and 1.875 MHz.
The Palit GeForce RTX 2060 Super JS consistently clocks higher, even though the power limit provides an identical 215 watts. The PCB probably works a bit more efficiently.
Nevertheless, the graphics card for Battlefield 5 (V) mostly clocks down. In Anno 1800 and Battlefield V, the full 1.935 MHz alternates at least with 1.920 MHz. The lowest clock is found in Hitman 2 and Metro Exodus with 1.905 MHz to 1.920 MHz.
If the power limit is manually maximized, both the Gigabyte and Palit graphics cards work with 1.935 MHz throughout.
On average, the Gigabyte GeForce RTX 2060 Super Gaming OC works 4 percent faster in Battlefield 5 (V)than Nvidia’s Founders Edition. Thus the graphics card overtakes a simple GeForce RTX 2070, while the GeForce RTX 2070 Super remains 17 percent faster.
In addition, the graphics card is 6 percent faster than AMD’s new Radeon RX 5700, but the Radeon RX 5700 XT remains 5 percent ahead. It is estimated that the graphics card is slightly slower than an inherently overclocked GeForce RTX 2070.
Cooling noise levels
The Gigabyte GeForce RTX 2060 Super Gaming OC is considerably louder. The fans tour with 2,170 revolutions per minute, which results in a measured value of 41.5 decibels.
Thus, the graphics card for Battlefield 5 (V) is 5 decibels louder than the GeForce RTX 2070 Super Gaming OC and it attracts attention with a decent noise when playing games, like for Battlefield 5 (V).
The graphics card isn’t disturbing yet, but the Palit product and its own bigger brother show that it could be quieter without any problems. On the Windows desktop, however, the 3D accelerator is silent.
Temperature problems are not expected on either 3D accelerator. The Gigabyte GeForce RTX 2060 Super Gaming OC heats up to 73 degrees Celsius, the Palit GeForce RTX 2060 Super JS stays 1 degree Celsius cooler. This also shows how much better the wide 2.5-slot cooler of the Gigabyte GeForce RTX 2060 Super Gaming OC is, because the graphics card stays 9 degrees cooler and is still quieter.
The TU106 GPU of the Gigabyte GeForce RTX 2060 Super Gaming OC can be increased by 123 MHz before the first crashes occur. The graphics card then clocks with around 2.040 MHz in Battlefield 5 (V). The 8 GB GDDR6 memory allows a plus of 904 MHz, which results in 7.904 MHz. Both measures improve the performance by about 7 percent.
In opposition to the Gigabyte GeForce RTX 2070 Super Gaming OC, the Gigabyte GeForce RTX 2060 Super Gaming OC can really convince. The manufacturer installs a smaller version of the Windforce cooler, which handles the requirements with bravour.
The graphics card remains silent under Windows, and its more silent than the Founders Edition when playing games like Battlefield 5 (V).
The additional 14 percent FPS are really worth it, which is why the Gigabyte GeForce RTX 2070 Super Gaming OC is ranking third on the best Graphics cards for Battlefield 5 / V.