Increase Hashrate GTX 1080 Ti GameRock Premium 11GB Mining: Overclock | MSI

How to ‘increase/ Max/ Optimize/ Boost/ Tune/ Improve’ Hashrate Palit GeForce GTX 1080 Ti GameRock Premium Mining + Gaming: Overclocking / Best Settings/ MSI Afterburner – Review and testing of the video card Palit GeForce GTX 1080 Ti GameRock Premium (PA-GTX1080Ti GameRock Premium 11G)

We continue our series of reviews of the latest GeForce GTX 1080 Ti series graphics cards. This time we got Palit ‘s flagship – GeForce GTX 1080 Ti GameRock Premium . This video adapter has the highest factory overclock in the company’s lineup, a redesigned LED backlight and presents us with its innovation in video card cooling systems – coaxial fans. Let’s check these features in action.

Increase Hashrate GTX 1080 Ti GameRock Premium 11GB Mining: Packaging and equipment

The packaging of the video card has not undergone any changes, except for the changed name of the video card.

The video card comes in a large box equipped with a carrying handle, which is rare. A trifle, but very convenient.

On the front of the box is the name of the video card, its amount of video memory and the NVIDIA technologies that this device supports.

The back of the box focuses on the new cooling system – TurboJET4 .

Despite the large size of the box, the internal content cannot be called extensive.

In addition to the video card in the antistatic package, there is only a software disk, instructions and an additional power adapter 2×6-pin -> 8-pin. The package bundle is quite poor, but this has already become the norm even for the flagships of manufacturing companies.

The main competitors of the video card

The main rival of the top graphics card on the market is the flagship RX Vega 64 from AMD. The cost of the model ranges from 33 to 50 thousand rubles, the clock frequency after overclocking reaches 1.54 gigahertz. A standard eight gigabyte memory module and low overclocking potential due to a weak cooling system reduce the profitability of the model.

The budget analogue of the mining farm board is NVIDIA Inno3D GeForce, a specialized module for cryptocurrency mining. The graphics card has no video outputs and is equipped with a five gigabyte RAM module. The processor speed after overclocking reaches 1.58 gigahertz, the cost of the GPU model is only 30 thousand rubles.

Design and features of the video card

The video card in question looks very similar to its predecessors in the GameRock series. The first difference that immediately catches your eye is the changed color scheme.

The reverse side of the video card has a backplate, which acts as a reinforcing and protective plate. The number of attachment points does not cast doubt on the quality of the assembly. But the presence of protective seals is somewhat upsetting.

The inside of the shroud that protrudes from the card and covers the heatpipes is now black and not as conspicuous as it used to be. At the top, next to the SLI connectors, is the BIOS switch. At the bottom of the far left corner, you can see the contact pad, which acts as a debugging panel for developers. Unfortunately, the contact pad with the points for measuring voltages and removing power and voltage limits on this video card has disappeared.

The physical dimensions of the video card have not changed: the length is still the same 285 mm, which ensures compatibility with most cases.

The video card occupies almost 3 expansion slots (58 mm), so if you want to install video cards in SLI mode, you should think about side airflow. On the front side there is also an illuminated inscription ” Palit “.

The rear interface panel has a good perforation, which helps to improve the temperature regime within the case. The panel has one Dual-Link DVI connector, three DP1.4 and one HDMI 2.0b.

Three large heat pipes peek out from the back.

There is also a connector for connecting RGB LED strips.

Overclocking technique for Nvidia graphics cards

In the AfterBurner program, we enable core voltage monitoring.

At 100% loading of cards (and equihash mining gives exactly constant 100%), the power consumption of the card depends on the voltage on the core.

It would be ideal to set the minimum voltage, but along with the voltage, the maximum overclocking frequency also drops sharply.

The dependence of the limiting frequency on the core voltage for this series is a typical exponential dependence.

Up to a certain voltage limit, frequencies can be increased seriously with each step, but after a certain maximum frequency, a further increase is almost useless.

We need to find a golden mean – when an increase in the output speed will be cost-effective in relation to the consumption of the card.

Of course, in conditions of super-profitability (May-June 2021 and December-January 2017-2018), it is generally unprofitable to limit the voltage greatly, but we are considering and developing systems that are planned to be used for months without switching. At the same time, a small lack of speed during the period of super profitability is more than compensated by serious savings during the period of medium and low profitability, lower energy bills, lower ventilation and cooling costs, noticeably higher density of installation of farms and long service life of video cards.

Cooling system and circuit board

Dismantling the cooling system is quite simple. It is necessary to unscrew the four screws around the GPU and disconnect the two connectors.

The new cooling system is called TurboJET4. Its dimensions, as well as the elaboration, are impressive.

Five heat pipes (two with a diameter of 6 mm and three more with a diameter of 10 mm) pierce the two sections of the radiator.

A huge copper base contacts the GPU, a plate soldered to the heatsink is located around it, which cools the memory chips through thermal pads.

To remove the casing from the cooling system, unscrew six screws.

The radiator fins are quite frequent, which should affect heat dissipation at low fan speeds.

Cooling of the power subsystem is performed qualitatively. The huge plate that is in contact with the main heatsink needs to efficiently transfer heat to the heatsink section.

With a casing and fans, everything is much more interesting.

The upper part is reserved for the LED backlight module, but instead of the usual fans, there are two coaxial ones.

The cooling system lives up to its name. Unfortunately, no fan markings were found.

A simple board with pin assignment is responsible for the synchronism of the rotation of the fans.

The rear reinforcement plate is no longer just for rigidity and protection. The backplate is equipped with a thin plastic strip through which LED light passes.

Also, the plate in the area of ​​the power subsystem is pressed against the back of the printed circuit board, which provides additional heat dissipation.

The printed circuit board of the video card is significantly different from the reference.

Twelve phases provide power to the GPU, two more – to the memory subsystem.

This circuit is controlled by NCP 81274 from On Semiconductor.

Additional power for the video card is made according to the 8 + 8 pin scheme. All this ensures high factory overclocking of the video card, and also does not exclude its further increase.

The video card has 11 GDDR5X memory chips marked 7BA77-D9VRL, which, together with a 352-bit bus, provide the video card with high bandwidth.

The reverse side of the printed circuit board is full of elements.

In the upper part there are 6 LEDs, which form the illumination of the back plate.

⇡#Performance: 3DMark

Synthetic tests ignore the loss of one gigabyte of memory in the GeForce GTX 1080 Ti, but are sensitive to GPU clock speed. As a result, in various tests, either the GTX 1080 Ti and TITAN X are equivalent in points, or the GTX 1080 Ti has a marginal advantage of 1-3%.

Whichever of the other single-socket gaming graphics cards we take to compare with the GTX 1080 Ti, the latter demonstrates a completely different level of performance. The new flagship is on average 32% ahead of the GTX 1080 in 3DMark, and in the case of the GeForce GTX 980 Ti and Radeon R9 Fury X, the advantage of the GTX 1080 Ti reaches 70% and 75%.

Performance: Gaming ( 1920×1080 , 2560×1440 , 3840×2160)

Tests in real games paint a more complex picture of comparing the GeForce GTX 1080 Ti and TITAN X. At all resolutions, the new product has an average advantage of 2%. At the same time, sometimes there is no difference even of 1 FPS between the two cards, and rare examples of TITAN X victory just come down to one frame per second in excess of the result of the GTX 1080 Ti, which clearly fits into random fluctuations in performance measurements. However, there is a game that clearly favors the GTX 1080 Ti – Tom Clancy’s The Division. In it, the advantage of the GTX 1080 Ti is from 8 to 10%.

Compared to the GTX 1080, the new flagship offers 25-33% faster frame rates, but the difference is greatest at higher resolutions, from 1440p to 2160p. The processing power of top gaming graphics cards is so great that in 1080p mode, CPU speed becomes an important factor. In addition, DOOM is already hitting the game’s 200 FPS frame rate cap on both GPUs.

The advantage of the GTX 1080 Ti over the GTX TITAN X based on the Maxwell architecture in games is smaller than in the “synthetics”, but is still extremely large – from 54 to 67% depending on the resolution. The Radeon R9 Fury X, on the contrary, turned out to be an even easier opponent, because a small amount of RAM (4 GB) plays against the AMD accelerator. As a result, if in 1080p mode the GTX 1080 Ti provides a 70% higher frame rate, then in 1440p and 2160p the difference increases to 82 and 96%.

We will pay special attention to the results of the GTX 1080 Ti in 4K. Only NVIDIA graphics cards based on GP102 are currently unconditionally suitable for playing at this resolution with maximum graphics quality settings (albeit without resource-intensive full-screen anti-aliasing). In most of the games from our test clip, the average frame rate exceeds 50-60 FPS, and only in Deus Ex: Mankind Divided does it fall significantly below this level.

Performance: video decoding

Chips of the Pascal family are equipped with the same multimedia block, the performance of which depends solely on the GPU clock frequency. As a result, there are practically no significant differences between the top “green” video cards in terms of the speed of decoding H.264 and HEVC video formats. In this discipline, Pascal has moved far ahead of past graphics cards, especially the GeForce GTX TITAN X, which, unlike the GTX 960 and 950, cannot fully decode HEVC in hardware and shifts some of the work to the CPU. AMD also, although it has improved the performance of its codec in H.264, cannot compete on equal terms with Pascal.

Note: since decoders usually do not differ within the same GPU family, the diagrams show one device from each family (or more if this rule is violated).

Software and lighting

The GTX 1080 Ti GameRock Premium graphics card is supported by Palit’s ThunderMaster software.

The program has not undergone visual changes, unfortunately. But its functionality is not inferior to the capabilities of a competitor.

There are monitoring of operating frequencies, temperatures and fan speeds of the cooling system, tools for overclocking the video card (the only thing, at the time of the review, the Master Mode tab was not available) and fan and backlight control, as well as saving and loading custom and preset profiles.

OC, Gaming and Silent modes differ only in base and boost frequencies.

The default mode is OC.

The backlight of the video card consists of several zones that have a common control of color and mode of operation.

Depending on the selected mode, the color may change depending on the temperature of the GPU: be static (color selection accuracy leaves much to be desired) or change over time (Rainbow). There is also a transfer mode to the NVIDIA panel.

The following video shows the most interesting mode of operation – Rainbow.

Hashrate when using various mining algorithms

The $1200 card was launched at 250W. It’s worth noting that the Founders Edition allows you to exceed the maximum threshold by 20% by running the card in 300W mode. But this will require good cooling so that your expensive purchase does not burn out.

All tests were performed without overclocking and with standard card settings, the cooler speed was set to maximum. The ccMiner 2.0 RC2 program was used, with the same standard settings. All mining algorithms supported by the miner were checked.

As a result, Blakecoin and Blake2s performed best of all, where the hash rate was 7521 and 6178 MH/s, respectively. The TDP was 100%. Hashing power and power consumption can vary greatly from algorithm to algorithm. But it is worth mentioning that algorithms that can overcome the stock TDP limit can produce much better results.

If you look at the long term, it is better to slightly reduce the load and work not at the limit of the TDP of the video card. After all, this solution will save energy.

The hashrate will not drop much, and the video card will last many times longer. To compensate for the slight drop in hashrate, you can connect a cheaper GPU with overclocking, in SLI, to the GTX 1080 Ti.

The video card is the most expensive solution from the entire manufacturer’s line, but at the same time it gives the best performance. Compared to the closest analogue, you can mine on 1080 Ti almost twice as fast as on 1070.

Here the question arises, why overpay and take the Ti-version if the standard 1080 looks just as good? The Ti gives about 700 Sol, while the standard 1080 gives 510-520.

At the same time, the difference in the cost of the cards is on average $100-200, and in the output power is almost one and a half times.

Specifications

The video card is based on the already familiar GP102-350-K1-A1, manufactured in the 10th week of 2021 in Taiwan.

Using the TechPowerUp GPU-Z program, we get some information of interest to us:

The most interesting thing: in addition to increasing the base and boost frequencies, this is an extension of the power limit to 350 W, which is enough for overclocking with an air cooling system without resetting the frequency.

NVSMI confirms this testimony.

New king! Testing GeForce GTX 1080 Ti

NVIDIA continues to conquer new heights of graphics performance. AMD still does not mature with the release of Radeon Vega for the segment of older solutions that could compete with last year’s “green” flagships, and now the new NVIDIA video adapters are in a hurry to empty the wallets of enthusiasts. The latest high-profile premieres were video accelerators GeForce GTX 1080 Ti and Titan Xp. The second is an exclusive product with a small circulation and an extremely high price. But the GeForce GTX 1080 Ti will be the new flagship of mass production models, which significantly raises the performance bar. We will talk about it in this article.

The GeForce GTX 1080 Ti is based on the GP102 GPU based on the Pascal architecture. This chip previously became the basis for the super-powerful Titan X (Pascal) solution, and the even faster Titan Xp has recently been released. The GP102 processor has six graphics clusters of five SM multiprocessors instead of four clusters for the GP104 (GeForce GTX 1080). The internal structure of clusters is absolutely identical, it is discussed in detail in a separate review of the Pascal architecture.

In full configuration, the GP102 has 3840 CUDA stream processors and 240 texture units. The memory bandwidth is increased to 384 bits thanks to 12 memory controllers. The crystal is produced on the basis of a 16-nm process technology and has 12 billion transistors.

Perhaps due to such a complex structure, the first GP102-based Titan X received an incomplete configuration with 3584 cores and 224 texture units. The new serial flagship GeForce GTX 1080 Ti has a similar configuration. The newcomer also has 88 ROP blocks with 2.8 MB of L2 cache (GP104 has only 2 MB).

The flexible structure of the new memory system made it possible to achieve an unusual bus configuration with a total capacity of 352 bits, which is why the memory capacity of the video card is 11 GB. Thus, a certain advantage of the Titan X, which has 12 GB, remains. However, for the average player, even 11 GB is enough for the most modern games. At the same time, the memory of the GeForce GTX 1080 Ti operates at a record frequency of 11 GHz.

After the release of the GeForce GTX 1080 Ti, NVIDIA also introduced the full version of the GP102 in the face of the Titan Xp. The GeForce GTX 1080 Ti is identical to the Titan Xp in terms of frequency configuration. The base GPU value of 1480 MHz is ensured even in the most extreme operating conditions. GPU Boost technology increases the core frequency until the power and temperature limits are exceeded. The average Boost value is 1582 MHz, but the maximum is traditionally higher.

You can compare all current solutions based on GP102 with their predecessors using the table below.

The GeForce GTX 1080 Ti combines all the advantages of the Pascal architecture and supports technologies that are also relevant for the GeForce GTX 1080. Among them, high bandwidth efficiency due to advanced frame buffer data compression methods. Also, tile caching gives a certain advantage, when the image is divided into fragments (tiles), which allows you to optimize the processing of geometry and pixels for each tile. This method allows more efficient use of L2 cache resources, reducing the load on external memory.

And all this is combined with a 352-bit bus and a record frequency of GDDR5X chips. With a physical throughput of 484 GB/s, the theoretical peak can reach 1.2 TB/s. However, the effectiveness of additional improvements is highly dependent on each specific application.

NVIDIA solutions are well-optimized for DirectX 11 environments, backed by years of support from engineers and programmers. Thanks to the availability of convenient tools, this API is still the main one for many game developers. DirectX 12 gives you more control over GPU resources, better use of CPU multi-threading capabilities, and more efficient combined operations, combining graphics and other computing, including thanks to Async Compute. NVIDIA Pascal generation solutions fully support all the features of the new API at the hardware level, including advanced asynchronous computing. But the introduction of DirectX 12 requires more investment on the part of developers, and the benefit from it is still small. Many games don’t get any performance benefit from DirectX 12 until several months after release after a series of updates. Examples of this are Rise of The Tomb Raider, Deus Ex: Mankind Divided and Seasonal Hitman. And these are good examples, because sometimes DirectX 12 brings only a drop in fps and nothing more.

NVIDIA is trying to improve the situation by providing software support for its products and introducing new developer tools for debugging code. Microsoft recently introduced the PIX utility for debugging under DirectX 12, and NVIDIA offers nSight Visual Studio Edition 5.3 and NVIDIA Aftermat. These programs will help DirectX 12 developers find and fix bugs.

The situation with software support is also constantly improving. Along with the launch of the GeForce GTX 1080 Ti, a new driver has been released that seriously speeds up a number of games under DirectX 12. And the improvement of the drivers made it possible to achieve noticeable acceleration in the already mentioned Rise of The Tomb Raider, Hitman and other games.

The company is also expanding its GameWorks effects experience to include support for the new DirectX 12. The initiative is based on a toolkit of off-the-shelf graphics technologies and effects that are easy to integrate and improve the visual performance of the game. NVIDIA GameWorks support is present in many modern games, some technologies are supported in the latest versions of Unreal Engine, Unity and Amazon Lumberyard.

The HBAO+ background shading technology has found wide application, which gives a better and more elaborate picture in relation to SSAO and other algorithms. Volumetric Lighting simulation technology proved to be excellent. This is exactly the lighting you saw in Fallout 4, and HBAO + support and Flex-based shard physics effects are already integrated into this game.

Another example of GameWorks is the HairWorks technology, which allows you to simulate complex hair and fur. The technology was actively used in The Witcher 3: Wild Hunt. Now HairWorks works under DirectX 12, and soon we may see new games with similar effects.

Among the new effects, Turf Effects should be highlighted for simulating complex grass with separate geometry of blades of grass.

The technology is used in Tom Clancy’s Ghost Recon: Wildlands. The game itself shows a great picture and supports many other NVIDIA technologies. Learn more about the technological capabilities of Ghost Recon: Wildlands with a demonstration of GameWorks effects in a special video:

Simulating various dynamic environments is a complex task that can be solved by different GameWorks tools. Real-time dynamically changing water can be implemented based on WaveWorks, Flex technology allows you to simulate the behavior of friable solids, Flow simulates flows and gases, Blast can be used for destructible objects, NvCloth will allow you to implement realistic fabrics.

Most techniques work under DirectX 12, including the promising Flex and Flow. As an example of the implementation of Flow, here is a short technical video demonstrating the new technology.

Optimization for the new API includes support for asynchronous calculations, which allows the GPU to more efficiently calculate such effects and get real acceleration on new video cards. And according to NVIDIA experts, Async Compute support provides a twofold acceleration of Flex effects rendering on the same GeForce GTX 1080.

The new GameWorks support DirectCompute and work on any graphics device. So you can play with the new effects on AMD graphics cards as well. However, at the present moment, almost all effects work on Radeon.

Neural networks in artificial intelligence systems have become a new area of ​​application for GPUs. In particular, NVIDIA is actively working in the field of artificial intelligence devices for unmanned vehicles. But there is a certain practical benefit for ordinary PC users. Based on self-learning neural networks, face recognition technologies and progressive image scaling methods are implemented.

The GeForce Experience software package is overgrown with new features, which allows you to automatically optimize game settings for your video card, update drivers and use additional features.

Part of these features is video capture with ShadowPlay. The built-in video block in the GPU is engaged in encoding, there is practically no performance drop in the game, the output video quality is very high, 4K resolution is supported. And the new ShadowPlay Highlights extension makes capturing highlights easier. The program automatically monitors the player’s activity and records the most intense game episodes.

Incredible opportunities gives players Ansel. The application allows you to take screenshots from unusual angles thanks to a free camera, apply color filters, create panoramic pictures and increase the resolution of the final image.

The features of Ansel appealed to the players. It is noted that users have already made more than 200 thousand screenshots in The Witcher 3 using Ansel.

In addition to this game, Ansel is supported in War Thunder, Wathc Dogs 2, Dishonored 2, Mass Effect: Andromeda and many other games. The SDK is available in the public domain, so any developer can integrate Ansel into their project.

Virtual reality VR is one of the rapidly developing areas. VR systems have found application not only in the entertainment industry, but also in specialized tasks. In an article on Pascal technologies, we wrote that special image projection techniques save GPU resources and speed up VR processing. NVIDIA VRWorks technologies provide VR rendering acceleration from 20% to 60% and more.

The combination of maximum performance and technology support makes the GeForce GTX 1080 Ti the best choice for those who want to touch the world of virtual reality.

After looking at common architectural features and technological innovations, it’s time to take a look at the reference design video card.

GeForce GTX 1080 Ti Founders Edition

The new NVIDIA flagship offers improved cooling and enhanced power supply compared to its comrades.

Externally, the GeForce GTX 1080 Ti differs from the GeForce GTX 1080 only by the name on the case. Reference versions of the Founders Edition are made in the same design.

The silver case is combined with black elements and large screws, which, together with numerous curves, creates a special “metallic” industrial style inherent in GeForce graphics cards. The total length of the adapter is 26.5 cm.

The reverse side is protected by a metal plate with an embossed pattern and the name of the video card.

On the side there is a large GeForce inscription with a green backlight. In the corner there are two power connectors – 6 and 8 pins.

As a result of design optimization, the DVI connector was abolished. This made it possible to free up space on the rear panel and make more holes for blowing air, slightly increasing the efficiency of the “turbine”. There are four interface connectors left: three DisplayPort and one HDMI.

The plate on the reverse side consists of two halves. Its main function is protective, but one of the plates is also in contact with the power controller on the surface of the textolite, thereby cooling it.

The overall layout of the cooler looks very familiar, similar to the GeForce GTX 1080 and earlier graphics cards. A large base covers the board, removing heat from all important components in the power circuit and memory chips. There is an additional heatsink at the edge of the board for better cooling of the power node. Removable black heatsink for graphics chip.

The GPU heatsink does not differ from the standard one in size and design.

At the base is a large evaporation chamber. It is claimed to improve its efficiency. Combined with a large number of holes on the rear panel, this can have a small positive effect on increasing the efficiency of a standard cooler. It should be noted that this type of cooling coped well with the GeForce GTX 980 Ti with its 250W TDP. And here, with the same values, it will definitely not be worse.

Let’s take a look at the PCB. In comparison with the GeForce GTX 1080, the circuitry is more complicated, there are more phases.

The core power subsystem has seven phases with two transistors per channel. The GDDR5X memory power node has two phases.

The full marking of the processor is GP102-350-K1-A1. In this case, recall, it operates with 3584 stream processors out of 3840 possible.

Soldered 11 Micron 6XA77 D9VRL memory chips with a total capacity of 11 GB.

Base core clock 1481 MHz with Boost Clock 1582 MHz. The data exchange rate of the memory corresponds to 11 GHz.

In terms of temperature and noise characteristics, the GeForce GTX 1080 Ti does not differ much from the GeForce GTX 1080. But if the younger video card has a temperature limit of 83 ° C, the Ti version has 84 ° C. When used in an open bench in a room with a temperature of 22 ° C, the fan speed under game load is about 2400 rpm, which is slightly higher compared to the GeForce GTX 1080. The noise is moderate.

Brief bursts of Boost at the moments of switching modes reach 1898 MHz, but the average Boost frequencies are closer to the level of 1700 MHz and below. In Deus Ex: Mankind Divided, the core frequency varied between 1683-1797 MHz, in Tom Clancy’s The Division it was about 1658-1749 MHz.

Among the test applications, only in Ghost Recon: Wildlands the average frequency level is consistently above 1747 MHz. Below is an example of the load when running the benchmark seven times in 4K with Very High settings.

It is worth noting that for such a powerful video card with a TDP of 250 W, the final frequency level will partly depend on the operating conditions. In a cramped case under severe thermal conditions, the temperature limit will be reached faster, resulting in a faster decrease in Boost. But since we are talking about a top product, it is assumed that the potential buyer will provide normal conditions and get the most out of his video card.

Overclocking allows you to further increase this maximum. The initial core frequencies of the GeForce GTX 1080 Ti are approximately on the same level as the GeForce GTX 1080. And what about overclocking? In the course of our experiments, we reached a level of about 2 GHz, losing very little in terms of frequencies to the forced GeForce GTX 1080 video adapter. We reached 1641 MHz at the base with a peak Boost value of up to 2050 MHz.

By increasing the power limit to the maximum (120%) and speeding up the blower fan, the average GPU frequency was maintained at 2000 MHz and below. So, in Deus Ex: Mankind Divided at 4K, the frequencies varied in the range of 1974-2025 MHz, and in The Division at 4K they even dropped to 1949-1962 MHz. The memory was overclocked to 6048 (12096) MHz. This is illustrated below.

In simpler modes and simpler games, the frequencies could exceed the 2 GHz bar, approaching the maximum. In any case, overclocking the GeForce GTX 1080 Ti requires powerful cooling. On a conventional video card, it is possible by manually accelerating the fan to an extremely noisy mode. This is suitable for testing the capabilities of the video card in 3DMark, but there is no need to talk about the practical application of such overclocking. This is possible on non-reference video cards with improved cooling. We will introduce you to some of these GeForce GTX 1080 Ti models in the near future.

The GeForce GTX 1080 Ti has no competitor from AMD. Therefore, our acquaintance with the new product will include testing in heavy modes and comparison with the GeForce GTX 1080, which occupied a senior position in the ranking of gaming graphics solutions for about a year.

Characteristics of the tested video cards

Let’s compare two reference models of NVIDIA video cards. Read more about the GeForce GTX 1080 in a separate article.

test bench

The test bench configuration is as follows:

• processor: Intel Core i7-6950X (3,[email protected],1 GHz);
• cooler: Noctua NH-D15 (two NF-A15 PWM fans, 140 mm, 1300 rpm);
• motherboard: MSI X99S MPower (Intel X99);
• memory: G.Skill F4-3200C14Q-32GTZ (4×8 GB, DDR4-3200, CL14-14-14-35);
• system disk: Intel SSD 520 Series 240GB (240 GB, SATA 6Gb/s);
• secondary drive: Hitachi HDS721010CLA332 (1 TB, SATA 3Gb/s, 7200 rpm);
• power supply: Seasonic SS-750KM (750 W);
• monitor: ASUS PB278Q (2560×1440, 27 ″);

• operating system: Windows 10 Pro x64;
• Driver: NVIDIA GeForce 381.65.

Testing was carried out at resolutions of 2560×1440 and 3840×2160 at maximum or high graphics settings.

Test Methodology

Battlefield 4

Testing was carried out in the first mission after the wall was blown up. Repeated running through a small area with dense vegetation before descending to a large construction site. Frame rate was measured using Fraps.

All graphics settings on Ultra, MSAA multisampling in 4x mode.

Battlefield 1

Testing was carried out by replaying a small game episode at the beginning of the Cape Helles mission, where the Gallipoli assault is played out. Soldiers land on the beach under heavy artillery fire, and with heavy explosions, performance sags more than in most other moments in the game. Seven repetitions were performed.

Ultra quality selected, video memory limit disabled. Testing done in DirectX 11 using Fraps.

Deus Ex: Mankind Divided

For testing, the built-in benchmark was used, which was run at least seven times.

Selected Ultra quality profile for 2560×1440 and Very High for 3840×2160. All tests were carried out in DirectX 12 mode.

Dishonored 2

Testing took place immediately after arriving in Karnaca, a walk was made around the port location. This is one of the hardest levels in the game, so the performance here is quite revealing.

Highest quality profile set to Ultra, all effects enabled, HBAO+ shading enabled. Disabled adaptive resolution, which degrades image quality in order to increase fps.

Fallout 4

Testing was conducted using Fraps immediately after leaving the hideout at the beginning of the game. There was a short walk through the surroundings with an abundance of vegetation and rich rays of light. Scenes with such environments result in the most noticeable performance degradations. The procedure is shown below.

The maximum graphics quality profile is selected. At lower resolutions, HBAO+ shading is additionally enabled.

Gears of War 4

The built-in gaming benchmark was used, which was run 6-7 times.

Ultra graphics quality is set to maximum, additional DirectX 12 features are active (Async Compute, etc.).

Grand Theft Auto 5

The built-in benchmark was used for testing. Five repetitions. For a comprehensive assessment, the average fps was calculated based on the results of all test scenes. The minimum fps was measured using Fraps following the results of a full pass of the benchmark.

All the main graphics settings are at maximum, additional parameters are active – the loading range of detailed objects (Extended Distance Scaling) and the “shadow length” item (Extended Shadows Distance) + 100% to the base level. At 2560×1440, MSAA 4x is enabled, at higher resolutions FXAA is active.

Titanfall 2

Testing was done with Fraps at the beginning of the first mission after the prologue. A certain sequence of actions was performed on a large location with a complex landscape. At least seven repetitions for each mode.

The maximum quality of all graphic parameters is selected, including the “insane” quality of textures. Additionally, adaptive resolution is disabled. At 2560×1440, heavy MSAA 4x was active, at 4K TSAA was used.

The Witcher 3: Wild Hunt

Testing was done with Fraps. Fps was measured during a trip on the road to the village of Bely Sad. At least six repetitions and an additional test session for the easiest graphics mode.

The maximum graphics settings are used, all post-processing effects and HBAO + shading are active. HairWorks enabled at 2560×1440, tested in 4K with it disabled.

Tom Clancy’s Ghost Recon: Wildlands

For testing, the built-in gaming benchmark was used, which was run seven times.

Presets selected for maximum Ultra quality for 2560×1440 and Very High for higher resolutions.

Tom Clancy’s The Division

The built-in performance test was run at least seven times for each mode.

The highest quality profile is selected. For 2560×1440, those parameters that were not initially set to this level were additionally increased to the maximum (quality of reflections, HBAO+ background shading, detail). At 3840×2160 we limited ourselves to the usual Ultra quality.

Watch Dogs 2

Testing in this game is slightly different from previous reviews. In this case, a walk was performed in the Palo Alto area with a run through the streets and a nearby grove with trees and dense grass.

The maximum graphics quality profile is set to Ultra at 2560×1440. For a higher resolution, similar graphics settings were combined with the activation of the “Temporal Filtering” option.

3DMark Fire Strike

Fire Strike benchmark from the latest 3DMark test suite. Testing was carried out in Extreme (2560×1440) and Ultra (3840×2160) modes.

3DMark Time Spy

New benchmark for DirectX 12, launched with default settings.

Energy consumption

The results of measurements in five appendices are given:

• Deus Ex: Mankind Divided;
• Gears of War 4;
• Grand Theft Auto 5;
• Tom Clancy’s Ghost Recon: Wildlands;
• Tom Clancy’s The Division.

The peak values ​​during each run were taken into account, on the basis of which the average value was calculated from the results of testing at both resolutions, and then the overall average was calculated. Data were taken using the Cost Control 3000 instrument.

Test results

Battlefield 4

We start studying the results with the earlier Battlefield 4. At a resolution of 2560 × 1440, even the GeForce GTX 1080 demonstrates an average frequency of 100 fps, the results of the Ti version are 35% higher. In 4K, the new flagship hits the 70fps milestone even at the initial frequencies, and this is in a heavy test scene.

Battlefield 1

In Battlefield 1, the advantage of the GeForce GTX 1080 Ti over its predecessor is more than 30%. Acceleration gives an additional acceleration of 13%. At 4K resolution, performance at 45-51 fps is not very impressive, but this is in an extremely difficult test episode. Often the frame rate is higher, and with a slight adjustment of the settings, you can easily reach more comfortable fps values.

Deus Ex: Mankind Divided

At 4K, there was a large variation in the minimum fps results. The final data here is heavily rounded, so they need to be compared taking into account the possible error.

In the Mankind Divided test, the difference between video adapters is at the level of 31–34% at initial frequencies. And if both comrades cope with the Ultra-mode at 2560 × 1440, then under 4K you already have to lower the graphics settings. At the same time, the GeForce GTX 1080 barely gives out about 30 fps in the benchmark, and the GeForce GTX 1080 Ti confidently overcomes the 40 fps mark. The built-in test simulates a very heavy scene, so in a real game, performance is higher most of the time. Therefore, there will be no serious discomfort in 4K on the new GeForce GTX 1080 Ti video adapter.

Dishonored 2

In the lower resolution of Dishonored 2, the difference between the rivals is at the level of 32-33%, when moving to 4K it is already 37-40%. The younger GeForce GTX 1080 does not compensate for such a gap with any overclocking. GeForce GTX 1080 Ti improves performance by 8-12% due to its increase in frequencies. It is worth noting that the game has wild requirements for video memory, loading almost the entire available video buffer. In the test scene, the GeForce GTX 1080 Ti loaded up to 9.2 GB of VRAM at maximum resolution, the GeForce GTX 1080 loaded below 8 GB.

Fallout 4

A serious gap between comrades and in Fallout 4. In lower resolution, when compared at standard frequencies, the new GeForce GTX 1080 Ti is faster by 33-37%, and in 4K the advantage is 37-40%. Increasing the frequencies provides an increase in results by about 10%.

Gears of War 4

Compared to previous games, the difference between video adapters in Gears of War 4 is not so significant. Perhaps the potential of a beginner is holding back the CPU. At a lower resolution, the difference between the cards is at the level of 15-25%, and with an increase in resolution it reaches 28%. The results in the second mode are impressive, and overclocking even allows you to get 60 fps on the GeForce GTX 1080 Ti. The game was able to load up to 7 GB of VRAM on NVIDIA’s flagship solution.

Grand Theft Auto 5

According to the results of testing in GTA 5, the advantage of the GeForce GTX 1080 Ti is less than 28%. At the same time, it is clearly seen that the simple GeForce GTX 1080 is in a different category, and no amount of overclocking can compensate for its lag. The performance of the Ti-version in 4K with simple FXAA is very high, more resource-intensive MSAA options can be enabled on this video adapter.

Titanfall 2

The GeForce GTX 1080 Ti performs well in Titanfall 2 at both resolutions. In the 2560×1440 mode, the advantage over its predecessor is up to 25%, with an increase in resolution it is a stable 29%. In simple mode, overclocking gives an acceleration of less than 7%, in more severe mode, more than 10%.

The Witcher 3: Wild Hunt

Ever dreamed of playing The Witcher 3 at 3840×2160 resolution? This is now possible on the GeForce GTX 1080 Ti, which delivers over 50 fps in Ultra mode. Enabling HairWorks can give some drawdowns, but fine-tuning the quality of the hair and additional post-processing options compensates for this. By the way, all this is discussed in detail in our special article about the nuances of the graphic settings of this game. At 4K, the difference between the GeForce GTX 1080 Ti and GeForce GTX 1080 is 35-36%, and at lower resolutions it’s about 30%.

Tom Clancy’s Ghost Recon: Wildlands

This new open world shooter has extremely high and serious system requirements. Therefore, for 4K, I had to limit myself to not the highest graphics settings. The difference between video adapters is up to 27% in lower resolution and 30-33% in heavier mode. Overclocking speeds up the GeForce GTX 1080 Ti by about 12%. Read more about performance in Ghost Recon: Wildlands on simpler graphics cards in a separate article.

Tom Clancy’s The Division

In The Division, we see a solid 35% advantage in average game frame rates for the GeForce GTX 1080 Ti. In terms of minimum fps, the difference is even greater, but due to the serious spread and possible error, we evaluate these indicators with caution. Both video adapters get about the same acceleration when overclocked, which is equivalent to an increase of up to 13% relative to the initial results.

Watch Dogs 2

In both test modes of Watch Dogs 2, there is a similar ratio between rivals with a difference of about 31.5%. The GeForce GTX 1080 Ti easily handles WQHD mode and shows good fps in 4K. Note that indicators in the second mode are possible only when “Temporal filtering” is enabled, which uses certain algorithms to speed up rendering. This gives a good increase in performance with minimal artifacts and loss in image quality. Video memory usage peaked at 6.4 GB during testing, with long gaming sessions it will be even higher.

3DMark Fire Strike

Let’s finish testing with a selection of results in different 3DMark tests.

In 3DMark Fire Strike, the GeForce GTX 1080 Ti has a 32-33% advantage over its simpler comrade NVIDIA. When comparing overclocked video adapters, the gap is about the same.

3DMark Time Spy

In the Time Spy benchmark under DirectX 12, there is a familiar gap of 33%. Overclocking speeds up the GeForce GTX 1080 Ti by another 10%, and the GeForce GTX 1080 improves its performance by 11% with an increase in frequencies.

Energy consumption

The new GeForce GTX 1080 Ti is also seriously different in terms of power consumption, significantly surpassing even the boosted GeForce GTX 1080 in terms of performance. Overclocking is also associated with a noticeable increase in power consumption, although it gives a slight increase in performance.

conclusions

With the release of the GeForce GTX 1080 Ti, NVIDIA has cemented its position as the leader in the graphics market. The competitor will never release its counterpart for last year’s king of gaming graphics, and here is a completely new solution with an impressive increase in performance, which can play modern games at 4K resolution. A serious advantage of the GeForce GTX 1080 Ti over the GeForce GTX 1080 is provided by the new GP102 graphics chip with 3584 stream processors. At the same time, the memory bandwidth has been increased – the 352-bit bus is combined with the record frequencies of the GDDR5X microcircuits. All of this makes the GeForce GTX 1080 Ti the fastest gaming graphics accelerator that delivers the highest performance possible. An alternative can only be Titan Xp,

According to the results of testing, the advantage of the GeForce GTX 1080 Ti over the GeForce GTX 1080 is at the level of 25-40%. But it should be noted that the maximum advantage is noticeable in the highest 4K resolution. In simpler modes, we sometimes run into limitations due to the performance of other PC components. And if we focus on the heaviest games and 4K mode, then we can safely talk about a stable lead of 33-35% with a maximum of 40%.

All this is combined with an equally serious increase in energy consumption and heat dissipation. The reference Founders Edition cooler does a good job of cooling with moderate noise, with a core temperature limit of 84 °C. In terms of their temperature and noise characteristics, such video cards are quite consistent with the flagships of the past and the year before last. Such models allow you to comfortably play at maximum performance, but only with standard frequencies.

If you want to squeeze even more out of the GeForce GTX 1080 Ti, this will require a serious improvement in cooling. Therefore, those who like to overclock graphics cards should immediately look at non-reference versions with powerful coolers. We will consider some of them in the near future.

Along with the development of technical devices, NVIDIA is also improving software support. Improved drivers allow some games to realize the very performance advantage of DirectX 12, which was talked about a lot, but it was not at the start. That is why we are slowly switching some test applications to DirectX 12 mode. The set of effects that can be implemented in games using NVIDIA GameWorks tools is expanding. The functionality of the software for users is also expanding. Ansel gives you incredible opportunities to take beautiful screenshots, and ShadowPlay will get more opportunities to automatically capture video of spectacular game moments.

test bench

• Motherboard: ASUS X99-A II (BIOS 1701);
• Processor: Intel Core i7-5960X Extreme Edition “Haswell-E” 3000 MHz overclocked to 4600 MHz at 1.28 V;
• Cooling system: self-assembled cooling system;
• Thermal interface: Arctic Cooling MX-4;
• RAM: 4 x 8GB KFA2 HOF Hall Of Fame 3200MHz(HOF4CXLBS4000M19SF162K; 14-14-14-28 CR1);
• Video card: Palit GeForce GTX 1080 Ti GameRock Premium / Nvidia GeForce GTX 1080 Ti 11 GB GDDR5x
• Power supply: Corsair RM1000i 1000 watts;
• System storage: SSD Samsung 850 Evo 250GB;
• Корпус: Thermaltake Core X5 Riing Edition.

Software:

• Operating system: Windows 10 x64 “Professional” with all current updates from Windows Update;
• Video card driver: Nvidia GeForce Game Ready Driver WHQL 382.33.

GeForce GTX1080 Ti under mining

If you have been mining Ethereum, then you probably know that AMD video cards are better suited for this. Geforce GTX 1080 video cards do not show high results when mining Ethereum and, moreover, at the moment they have some problems when mining under Windows.

Geforce GTX 1080 video cards use high-performance GDDR5X memory, which is very good in games, but apparently not very suitable for mining cryptocurrencies on algorithms that intensively load video memory.

Actually Geforce GTX 1080 is slower in Ethereum mining due to the use of GDDR5X video memory, even Geforce GTX 1070 with regular GDDR5 video memory has a higher hash rate on the Dagger-Hashimoto (Ethereum) algorithm. And given the fact that Geforce GTX 1080 is much more expensive than Geforce GTX 1070, it is not the best option for ETH mining.

We compiled the binaries from the latest Genoil’s ethminer 0.9.41 fork version 1.1.3 source code (source code) using CUDA 8.0 and Compute 6.1, which supports mining on the new Geforce GTX 1080 and Geforce GTX 1070 graphics cards and which you can download from the link below.

If you are using Windows to mine Ethereum, immediately after starting the miner with default settings, you will encounter certain problems. It will fail when trying to load a DAG file into the Geforce GTX 1080 video memory, no matter if you use CUDA or OpenCL.

Other OpenCL miners, such as qtminer, will also not start, giving a video driver failure. And even if you manage to run the miner, you will get depressingly low performance.

You can run Genoil CUDA fork of ethminer in CUDA mode with -U option and add following parameters to .bat file “–cuda-grid-size 2048 –cuda-block-size 128” to prevent driver error, however you will be able to get hashrate that will not exceed 1 Mhs / sec – which is very, very small for a video card of this level.

Cooling system testing and overclocking

To test the video card with factory and stable game overclocking, applications from the Futuremark 3DMark package were used. The room temperature was 25 °C, the case was open, there was no additional airflow for the video card.

In the Time Spy stability test, the graphics card at factory settings showed the following results:

As you can see in the graph, the video card in the factory mode constantly ran into power limits and, as a result, dropped frequencies and voltages in order to keep consumption within 250 watts. In this mode, the video card warmed up to a maximum of 70 ° C at 1400 rpm, and the frequencies did not exceed 1936 MHz.

After increasing the power limits alone to the maximum (350 W), the picture changed dramatically.

The GPU frequency increased to 1974 MHz, consumption did not exceed 120% out of 140% possible (300 W out of 350 W).

The temperature of the GPU increased by only 2°C and did not exceed 72°C, the fan speed increased to 1500 per minute.

In this mode, the video card submitted the following results:

There were no problems with overclocking. Without changing the fan speed control algorithm and the voltage of the video card, we managed to achieve a frequency of 2050 MHz for the graphics processor and 1514 MHz (12112 MHz effective frequency) for memory chips.

The fan speed remained practically unchanged, barely exceeding 1500 rpm (at the same time, the noise from the video card was comfortable), and the temperature rose only to 74°C. The result is good, but from such a powerful cooling system, according to the experience of previous video cards in the series, we expected a little more.

At the same time, the results in 3DMark grew slightly – by only 4%

Whether these few percent increase in consumption and temperature of the video card are worth it is up to everyone to decide. We believe that for video cards of this performance level, it makes more sense to undervolt by reducing consumption and temperatures without changing the performance level.

Downclock

230 watts:

190 watts:

180 watts:

165 watts:

Unfortunately, AfterBurner does not allow editing points below 800 mV (although you can go there by reducing the PL even lower). This does not interfere much, because, as we see, the optimal ratio is higher.

In our opinion, it is optimal to have a speed of 690-700 sols at 190 watts.

Yes, an increase of 60 watts will add 70 sols (60W for 70 sols – 2 times worse than mining on old AMD 7th generation cards), is this a logical ratio?

IMPORTANT! All other points from the found ones are set equal in frequencies to the found ones.

We get STAGES:

We can already walk these steps manually, decreasing or increasing the power limit (PL) parameter.

It is interesting to do this in automatic mode, more on that below.

IMPORTANT! Often, Tishka (as the very expensive and powerful 1080ti is affectionately called) does not immediately save the points (especially under load) or changes them according to its internal algorithm. Because of this, you have to expose them several times.

To check the correctness of saving the points, we switch to another map, and then we return and check the points of the curve.

Gaming Performance

In games without a built-in test application, the minimum and average fps readings were recorded using the Fraps program with the following settings:

Game tests were carried out at resolutions of 1920×1080, 2560×1440 and 3840×2160 with maximum graphics settings, only anti-aliasing settings were changed depending on the resolution.

For Battlefield 1, Crysis 3, The Witcher 3 Wild Hunt and Watch Dogs 2, the following game segments were used (they were filmed on a GTX 1060 6G video card, and monitoring on them has nothing to do with the video card in question):

The following results were obtained:

Automatic switching PL

To manage and monitor the power limit at your own discretion, use the NVSMI utility from the map developer.

You can download it here – https://yadi.sk/d/bQ5uqzm13RV5hR

We put this folder, for example, in “c: Programs” and when using executable batch files we get information about the cards and manage them.

Changing power is very easy. For example, to set a limit of 250W for card #0, we need a file with the content:

C:ProgramsNVSMInvidia-smi.exe -i 0 -pl 250

Result:

But first we need a list.bat file

Its content:

nvidia-smi.exe -L pause

It lists the cards in the system. Now we can match the names of the cards to their numbers.

But we have the same cards! What to do?

We use the consumption display executable file (example s_power_0.bat) with the following content:

nvidia-smi -i 0 —loop-ms=1000 —format=csv,noheader —query-gpu=power.draw

It displays the current consumption of the card with the specified number, and so on for any card. By starting mining on only one card, we can easily understand what it is by number.

Having decided on the numbering of the card, we create common files for switching power on all cards at once.

В нашем случае мы создали 3 файла (hard, normal, soft), которые можно активировать для ночного тарифа, для тарифа полупик и для тарифа пик соответственно.

ВАЖНО! Смена PL возможна только при исполнении файла от имени администратора, поэтому в каждом файле прописаны полные пути к утилите NVSMI.

Идём в меню Пуск->Стандартные->Служебные->Планировщик заданий

В планировщике ставим запуск режима HARD после 23ч, режим SOFT в 7ч и в 17ч, режим NOrmal в 10ч и 21ч. Это соответствует времени трёхтарифного переключения в Москве, что позволит не пережигать впустую электроэнергию в час пик и получать максимум в дешёвом ночном тарифе.

Related posts:

Download Lexicon Alpha Driver Free Windows (10 &7) 64-bit bit Windows Drive is 100% charged in How to fix the problem GTX 1050 2GT VS RX 560 Hashrate: Overclocking | Settings (Shocking) How to prepare graphics card to calculating cryptocurrencies (whattomine) How to use the Favorites section in the TikTok app: Miracle Solution How to Increase HD 8570 2gb Mining: Overclock| Hashrate| PhoenixMiner 5.5C