Buildings What is a fortified area in world of tanks

Selection of the best non-premium tanks for farming silver in World of Tanks Wot what is the best way to farm silver

What is the NES file extension?

Review of free programs for viewing NFO files

GDB file extension My computer does not show file extensions, what should I do?

Arduino Uno WiFi: description, characteristics, connection diagram

Video cards. Review of the MSI N760 HAWK video card Study of the potential of the cooling system

In the second half of 2013, after the announcement of two top-end consumer video cards, Nvidia introduced a new product in the mid-price segment. It turned out to be Nvidia Geforce GTX 760 - a video card based on the Kepler architecture. Nvidia's recommended price is 14,000 rubles or $230. Let's see what this video adapter is capable of.

You should start with the technical parameters of the model. Detailed specifications of the GTX 760 are presented in the table below.

Technological process, nm	28
Number of transistors, billion	3.54
Processor frequency, MHz	980
Memory frequency	1502
Memory type	DDR5
Memory capacity, GB	4
DirectX version	11.0
Memory bandwidth, GB/s	192

The characteristics of the Nvidia Geforce 760 are quite enough to enjoy the game at medium/maximum settings. Buyers are offered a GTX 760 2 Gb or 4 Gb. For users who play modern games, the best option would be to buy a GTX 760 4 Gb. This will ensure good performance and make it possible to play with maximum quality textures and shaders.

Review of the GTX 760 video card

A detailed review of the GTX 760 will allow you to appreciate all its advantages. Looking at the characteristics of the video adapter, you can see that this model is an improved 660 Ti. For the reference model, Nvidia is using the same design that the 660 and 670 had with a small PCB. The video adapter has a centrifugal fan. The model turned out to be quite compact, it occupies two expansion slots and will fit into almost any case.

The standard cooling system in the reference model looks very simple: it consists of a small aluminum radiator, a 70 mm fan and a plastic casing. The fan is equipped with PWM control with a built-in system for changing the rotation speed, which was used in the 780 and 770 models. Adaptive temperature control allows you to change the fan speed smoothly, making the video card more acoustically comfortable.

The temperature of the video adapter without load is 33-35 degrees, and the fan operates at a rotation speed of 1000 rpm. Under load, the operating temperature of the Nvidia GTX 760 video card rises to 83 degrees, and the cooler rotation speed rises to 2000 rpm. The fan does not make much noise when under load.

The rear panel contains the following connectors:

1 x DVI-I;
1 x HDMI;
1 x DisplayPort.

It should be noted that the video card has excellent automatic overclocking performance. With a base GPU frequency of 980 MHz, it automatically accelerates to 1150 MHz, and is capable of operating in this mode for quite a long time. In resource-intensive applications that put the maximum load on the video adapter, GPU Boost 2.0 increases the core frequency to 1100 MHz.

What kind of power supply is needed for GTX 760

According to the technical specifications, the power consumption of the GTX 760 is 170 W. The manufacturer recommends using power supplies with a power of at least 500 W for this video card. A +12 V line is also required.

How to overclock an Nvidia Geforce GTX 760 video card

Recently, high-performance video cards have been used not only for games, but also for cryptocurrency mining. To increase mining income on the GTX 760 by 10-15%, it will need to be overclocked.

Overclocking the Nvidia Geforce GTX760 video card can be done using MSI Afterburner. The algorithm of actions is as follows:

Download the utility from the official MSI website.
In the main window, increase the “Core Clock” and “Memory Clock” parameters by about 5-10 percent.
Test the video card in games for artifacts and video driver failures.
If the video adapter driver reboots or artifacts are visible on the screen, slightly reduce the settings.
If nothing happens, increase the parameter values a little more.

If you use the GTX 760 for mining, you will need good cooling. In this case, you need to slightly move the “Fan Speed” slider, which is responsible for the fan rotation speed. If you follow these rules, overclocking the GTX 760 will be completely safe.

On average, the processor frequency of the video adapter can be overclocked by 15% (up to 1120 MHz); in resource-intensive applications, the automatic overclocking mechanism accelerates it to 1250 MHz. The memory frequency can be overclocked by 17% (to 1700 MHz), which is a common result for the GK104 GPU.

Game testing results

Let's move on to testing the video card in modern games. The GTX 760 test was carried out at maximum graphics settings and a resolution of 1920x1080. Anisotropic filtering and MSAA 4x anti-aliasing are also included.

A game	Number of frames per second
Lost Planet 2	73
Sniper Elite 2	43
Far Cry 3	38
Crysis 2	35
Batman Arkham City	62
Metro 2033	31
Max Payne 3	42
Sleeping Dogs	61
CoD: Black Ops 3	77
GTA 5	54

To enjoy high FPS in modern games, users will have to sacrifice picture quality and lower anti-aliasing and shadow settings. Of course, to provide good frames per second, you will also need a good processor. A good option would be fifth-generation Intel Core i3 processors, such as the 5157U.

The video card is suitable for those users who use monitors with a resolution no higher than 1920x1080 or 1920x1200 pixels. It is at this resolution that the video card will be able to provide acceptable FPS at maximum settings.

Comparison of manufacturers

Since the difference in the characteristics of video cards from different brands has virtually no effect on performance, the main factor that guides users is how much the Nvidia Geforce GTX 760 costs. A comparison of different manufacturers is presented in the table below.

Parameter	Gigabyte GTX 760 2 Gb windforce 3x	Palit Geforce GTX 760 jetstream 2 Gb	Asus	Msi	Zotac
Memory frequency, MHz	1502	1550	1502	1502	1550
Core frequency, MHz	1085	1072	980	1033	1033
Bandwidth, Gb/s	192.3	198.4	192.3	199.4	196.4
price, rub.	17000	18500	14000	16500	16000

The price for Nvidia Geforce GTX 760 varies from 14 to 18.5 thousand rubles. Users who want to slightly increase performance are advised to purchase a video adapter from Palit or pay attention to more powerful models in the Nvidia line.

Latest drivers

The software is included with the video adapter. If you have lost the disk or want to update the Nvidia Geforce GTX 760 drivers to a new version, go to the corresponding tab on the company's official website. After that, select your video card and operating system (for example, Windows 7 64 bit) and download the appropriate file.

If you have any difficulties, you can download drivers for Nvidia Geforce GTX 760 in another way. To do this, click the “Graphics Drivers” button, after which suitable drivers will be automatically detected.

Everything you need to know about “free” FPS in games.

It's already the end of summer, and video cards are in no hurry to reduce their prices. The annual autumn wave of releases will begin very soon, and many of them will require powerful hardware responsible for graphics. Try to follow our advice so as not to be left behind in the new gaming year.

Note: This guide applies exclusively to desktop PCs and is devoted to the so-called “safe overclocking” (without raising the voltage or changing the cooling system). However, with any manipulation of the hardware, there is always some chance of damaging it. Therefore, read our advice very carefully and follow it at your own peril and risk.

Step 1: Prepare the computer

If, after opening the side wall of the system unit, you feel like sneezing at best, then quickly arm yourself with a rag and a basin of warm water. After first unplugging your gaming machine, removing your sweater and moving the woolen rug off the floor, thoroughly wipe the inside of it and blow out all the dust from the filters (it’s best to do this outside). It would also be good to change the thermal paste under the video card cooling system.

The second thing you need to do is make sure that you have sufficient wattage from the power supply. To do this, find the model of your graphics card (for example, GeForce GTX 760 or Radeon R9 280X) on the website of any major store and pay attention to the “Recommended power supply” column. If it says “400 W” and exactly that much (or even less) is available on board your PC, then after overclocking, freezes and even system shutdowns are possible - higher frequencies always increase power consumption.

As for the power supply parameters, it is worth setting either balanced or high performance mode. Better balanced - it will allow the video card to “rest” when the game is not running. Contrary to popular myth, it does not affect productivity, but rather uses resources wisely. You should not turn on, perhaps, the economy mode.

Step 2: Install the necessary software

The most important programs for any novice overclocker are GPU-Z and MSI Afterburner. The first one (in the image on the left) displays all the characteristics of your video card: from the exact model name (Name column) and core (GPU) to frequencies (Clock), memory bus width (Bus Width) and support for various technologies (Computing). The second allows you to increase the core (Core Clock) and memory (Memory Clock) frequencies, as well as control the behavior of the cooler (Fan Speed) and play with the voltage (Core Voltage). The latter, by the way, is not recommended for safe overclocking.

Any software, of course, should be downloaded only from official websites, so as not to run into various “surprises” when installing it.

Superposition Benchmark from Unigine

Among other things, it is also worth installing one of the popular performance tests from Unigine: Heaven Benchmark, Valley Benchmark, or the completely new Superposition Benchmark. If your graphics card is more than five years old, then the ultra-modern Superposition may not start - then install any of the first two.

Benchmarks allow you not only to admire three-dimensional scenes that are almost as beautiful as modern games, but also to quickly “warm up” the video card after overclocking in order to check the system’s stability.

Step 3: Test current performance

First, run GPU-Z and compare the values from the Default Clock (standard graphics core frequency) and Boost from the top line (the current graphics core frequency in games). If they are different, then your video card is factory overclocked and the frequency increase in Afterburner will be added on top.

Next, look at the number of megahertz in parallel graphs: GPU Clock and Default Clock, as well as in the upper and lower Memory. At the top are the current frequencies without taking into account factory overclocking, and at the bottom are the original ones. Paired items must match each other. Otherwise, your graphics card is already overclocked. Later we will definitely reset it through the MSI Afterburner program.

Now run the built-in performance test in any demanding game (for example, GTA 5 or Rise of the Tomb Raider) and record the average frame rate in a notepad. After that, open one of the Unigine benchmarks mentioned above, do not change anything in the settings (they will adjust themselves to your computer) and click the Run button.

Upon completion of testing, the benchmark will display the results: total points, minimum, maximum and average FPS, maximum temperature of the graphics card, and so on. All this data should also be written down.

Step 4: Raise the frequencies

To work with video card frequencies, you will need the previously downloaded MSI Afterburner utility. Right after launch, it will most likely look different from the image below. Nothing is clear about the new interface, so just go to settings, go to the “Interface” tab and select Default MSI Afterburner v3 Skin. Much better now.

For safe overclocking, we only need two sliders: Core Clock and Power Limit. The first adds the number of megahertz to the current core frequency in games (in GPU-Z this was the Boost item), and the second increases the power consumption limit of the graphics card. By default, MSI Afterburner is configured in such a way that Power Limit can be turned up to maximum and this will only add stability (but a power supply with a reserve is required). As for the Memory Clock item (current video memory frequency), it is better not to touch it at all - on graphics cards of recent years there has long been no benefit from overclocking video memory, only stability problems.

First, just in case, press the Reset button (in case the video card is already overclocked). Now you can add a few megahertz to the Core Clock column, but not more than 100 units. After that, click the Apply button and run the benchmark again. In extreme cases, the test utility will simply turn off during operation - which means you need to reduce ten megahertz and try again. If the benchmark reaches the results, then do not save anything yet and gradually move towards increasing the frequency until the benchmark begins to turn off. As soon as this happens, return to the previous Core Clock value.

Once a stable core frequency has been found, all that remains is to find out how much performance has increased. Everything is simple here: we run the benchmark for the last time, and then the built-in performance test from GTA 5 or other games mentioned above. Now we compare the results after overclocking with those recorded in step 3.

If the increase is at least 10%, then this is already very good (the maximum with air cooling is about 15%). In this case, you need to open MSI Afterburner, press the Save button, then the flashing number 1 and lock the lock on the left. Now overclocking can be quickly activated using this number.

If, after overclocking, only 1-2 frames are added in games, then it may be better not to risk it and press the Reset button.

Such good temperature performance is ensured thanks to high fan speeds, which spin up to 2500 rpm. At such a significant speed one would expect a high noise level, but this is not entirely true. We have repeatedly noted the low noise of FirstD fans of ASUS video cards. And even at such a high speed they do not create noticeable discomfort. The nature of the noise resembles a rustle, which is more pleasant than the hum of the Palit JetStream, although the latter's speed is supposedly lower.

In normal games, the video card turned out to be even quieter. For example, the game GRID 2 warmed the core up to only 67 °C at a maximum fan speed of 2400 rpm.

When overclocking, we managed to raise the base frequency of the GPU by 110 MHz at a standard voltage of 1.2 V. The available softvolt mode of 0.012 V did not bring any practical benefit.

The actual frequency in Boost reached 1267 MHz. This is more than Palit's modest overclock, but slightly less than the frequency we achieved on the MSI. The memory worked stably at 7012 MHz. This is better than memory overclocking from the same Palit, which is based on a reference PCB with two-phase GDDR5 power supply.
To avoid overheating, fan speeds were increased to almost 3000 rpm. Characteristics of tested video cards

We will compare the reviewed video adapter with the GeForce GTX 670, which will be played by ASUS GTX670-DCMOC-2GD5. The standard GeForce GTX 760 will be the ASUS GTX760-DC2OC-2GD5 with a fixed core frequency of 1097 MHz, which is quite consistent with the average value of the aggressive Boost of the reference samples. In previous articles, we compared the new GeForce only with older graphics accelerators. This time the GeForce GTX 660 Ti was added to testing, but only at nominal value. An extended comparison of the GeForce GTX 760 and GeForce GTX 660 Ti, taking into account the overclocking potential of each, will be presented in one of the future reviews. Radeon HD 7970 and Radeon HD 7970 GHz Edition have been added from AMD. Since these video adapters are from the higher price range, in this article we will limit ourselves to the results without overclocking. We also note that our GHz Edition is derived from a simple Radeon HD 7970 by setting the Boost mode to a fixed frequency. That is, instead of an option with a floating frequency in the range of 1000-1050, we have a video adapter with a constant value of 1050 MHz.

The technical data of the test participants is given in the table below. For GeForce, official Boost Clock data is provided. The graphs show the full range of core frequencies, including short-term peak values.

Video adapter	GeForce GTX 670		GeForce GTX 760	GeForce GTX 660 Ti	Radeon HD 7970 GHz Edition	Radeon HD 7970
Core	GK104	GK104	GK104	GK104	Tahiti	Tahiti
Number of transistors, million pieces	3500	3500	3500	3500	4312	4312
Technical process, nm	28	28	28	28	28	28
Core area, sq. mm	294	294	294	294	365	365
Number of stream processors	1344	1152	1152	1344	2048	2048
Number of texture blocks	112	96	96	112	128	128
Number of rendering units	32	32	32	24	32	32
Core frequency, MHz	915-980	1006-1072	980-1033	915-980	1000-1050	925
Memory bus, bit	256	256	256	192	384	384
Memory type	GDDR5	GDDR5	GDDR5	GDDR5	GDDR5	GDDR5
Memory frequency, MHz	6008	6008	6008	6008	6000	5500
Memory capacity, MB	2048	2048	2048	2048	3072	3072
Supported DirectX Version	11.1	11.1	11.1	11.1	11.1	11.1
Interface	PCI-E 3.0	PCI-E 3.0	PCI-E 3.0	PCI-E 3.0	PCI-E 3.0	PCI-E 3.0
Declared TDP level, W	170	170	170	150	250	250

Test bench

The test bench configuration is as follows:

processor: Intel Core i7-3930K ([email protected] GHz, 12 MB);
cooler: Thermalright Venomous X;
motherboard: ASUS Rampage IV Formula/Battlefield 3 (Intel X79 Express);
memory: Kingston KHX2133C11D3K4/16GX (4x4 GB, DDR3-2133@1866 MHz, 10-11-10-28-1T);
system disk: Intel SSD 520 Series 240GB (240 GB, SATA 6Gb/s);
additional drive: Hitachi HDS721010CLA332 (1 TB, SATA 3Gb/s, 7200 rpm);
power supply: Seasonic SS-750KM (750 W);
monitor: ASUS PB278Q (2560x1440, 27”);

operating system: Windows 7 Ultimate SP1 x64;
GeForce driver: NVIDIA GeForce 326.41;
Radeon driver: ATI Catalyst 13.8 beta.

User Account Control, Superfetch and visual interface effects are disabled in the operating system. Driver settings are standard, no changes. All tests were carried out in 1920x1080 resolution with maximum graphics settings. All details and nuances are described below.

Testing methodology

Frame rate was measured using Fraps. The mission “Rock and a hard place” has been selected. After the first checkpoint we go down the hillside into the valley to the enemy fortifications. The test scene included, in addition to the descent, the beginning of a firefight during the assault on the first row of fortifications. The fire was fired at two points through a sight from behind the stones to the right and left of the road. Taking into account a simple descent, the general procedure is easily repeatable, and the final results are almost independent of random factors. Plus, the frame contains not only a large area with detailed textures, but also the lighting effects of shots and explosions. This helps to create a completely adequate picture of real-life performance, both in complex, intense scenes of the single-player campaign, and in multiplayer battles. All detail and quality parameters are on Ultra under DirectX 11, the highest possible anti-aliasing mode AA4x is used.

Testing was carried out using the built-in gaming test. Three repetitions. Maximum graphics quality settings for DirectX 11 with active anti-aliasing.

The test was conducted at the beginning of the Only Human level. A short run through a limited visibility zone was combined with an inspection of distant objects through a sight (that’s why such a large-scale level was chosen). Four reps. Maximum graphics quality without anti-aliasing.

We took a walk through the territory of the first village and the surrounding thickets. The abundance of garlands and lights in the village creates a significant load on the video card, which is why this place was chosen for testing. Four reps. Maximum graphics settings for DirectX 11, HDAO, MSAA 4x anti-aliasing.

Five runs of the built-in performance test. All graphics settings at maximum, anti-aliasing MSAA 4x.

Four runs of a standard gaming benchmark test. Maximum image quality settings for DirectX 11 with MSAA 4x anti-aliasing (FXAA disabled).

As an experiment in this game, we decided to conduct testing, both using the built-in game test, and during a real playthrough by replaying a small episode at the Ashes level. A video of the corresponding scene is below.

Testing was carried out without activating enhanced physics effects and with Advanced PhysX enabled. In both modes, the graphics settings are set to maximum with the Motion Blur parameter set to the minimum value. Antialiasing is disabled. Note that due to minimal interaction with the environment at the Ashes level, the impact of PhysX on the results should be modest.

Unigine Valley benchmark

Ultra quality at 1920x1080 resolution and MSAA 4x anti-aliasing. Three repetitions.

remember me

Testing was carried out using Fraps. The opening scene for chapter seven has been selected. The duration of the test segment is 54 seconds. Graphics settings are set to High, SuperSamling is enabled, normal anti-aliasing is disabled. Five reps.

Tomb Raider

The built-in benchmark does not in any way characterize the actual performance in the game. It can only be considered as a demonstration of TressFX hair simulation technology. And this simulation itself is not used to its maximum in the test; there are also more complex scenes in the game when the hair is affected by strong wind or water. Therefore, we abandoned the built-in test and used the introductory scene at the very beginning of the game to compare performance: Lara is floundering in the waves, gets ashore and watches her comrades argue, after which an unknown person stuns the heroine. Four reruns of this episode. Testing in two modes: maximum graphics quality with FXAA anti-aliasing; similar settings when activating TressFX. All additional visual effects included.

Testing on the Extreme profile, which assumes the use of a resolution of 1920x1080.

Testing in normal mode at 1920x1080 resolution. Test results

The GeForce GTX 760 outperforms the GeForce GTX 660 Ti by 7-8% and is inferior to the older GeForce by about the same amount. The younger Radeon beats the hero of our review by only a couple of percent. But the accelerated ASUS is slightly ahead of the Radeon HD 7970 GHz Edition.

We see minimal difference between the regular GeForce GTX 760 and the GeForce GTX 660 Ti. ASUS is 4% more productive than the reference version of the GTX 760 and 5-6% weaker than the Radeon HD 7970. When overclocked, it manages to outperform the GeForce GTX 670. The souped-up ASUS does not reach the level of the Radeon HD 7970 GHz Edition.

The GeForce GTX 760 is 9% faster than the GeForce GTX 660 Ti and 7-8% weaker than the GeForce GTX 670. The difference between the ASUS GTX760-DC2OC-2GD5 and the reference is 4%. A simple GeForce shows a modest advantage over the younger Radeon, and an overclocked ASUS easily outperforms the older Radeon GHz Edition

The largest gap between the GeForce GTX 760 and the GeForce GTX 660 Ti is observed - 10% in terms of average fps and 20% in terms of the minimum parameter. Minimal difference between the reference and ASUS. Both cards are slightly inferior to the Radeon HD 7970. When operating frequencies are increased, the GTX760-DC2OC-2GD5 outperforms the older GeForce running at nominal and minimizes the lag behind the older Radeon in terms of minimum fps.

The game is based on the game engine DiRT: Showdown, which is known for its optimization for AMD solutions. So, quite predictably, GRID 2 demonstrates a significant advantage of Radeon video cards. The GeForce GTX 670 outperforms the younger Radeon only when overclocked. The difference between the GeForce GTX 760 and GeForce GTX 660 Ti is on average about 3%, which corresponds to ASUS's advantage over the reference. Overclocking further speeds up the hero of our review by 11-14%, which allows us to outperform the GeForce GTX 670 at standard frequencies.

The latest Hitman also prefers Radeon. Here the gap between AMD representatives is so significant that even the overclocked GeForce 670 does not reach the performance level of the Radeon HD 7970. The simple GeForce GTX 760 is 14% better than the GeForce GTX 660 Ti and 6-7% weaker than the older GeForce.

First, let's look at the results of the built-in benchmark.

Without PhysX, all NVIDIA participants are nominally inferior to the Radeon HD 7970. The difference between the GeForce GTX 660 Ti and the GeForce GTX 760 is 3%. The difference between the latter and the ASUS video adapter is 4%. Overclocking ASUS allows you to achieve a 15% advantage over the reference and makes it possible to win a couple of percent over the younger Radeon.

When activating advanced PhysX, NVIDIA representatives go ahead. The older Radeon is inferior even to the GeForce GTX 660 Ti.

Let's see what performance the participants will demonstrate in a real game at the Aches level.

Noteworthy is the slight drop in average fps when advanced physical effects are enabled. This is due to the fact that in the selected scene there are no shootouts or destructible objects. Firing a shotgun only renders more shrapnel and smoke. This is not enough to significantly reduce the average result. But this is quite enough to show how significant the drawdowns on the Radeon can be during shooting. And during mass battles, the situation for AMD representatives will be even worse, which is clearly shown by previous results in the benchmark, which simulates just such a scene at the D6 level. But even without “physics” in a real game, NVIDIA cards show good results precisely in terms of minimum fps - pay attention to the identical performance of the regular GeForce GTX 760 and Radeon HD 7970.

The GeForce GTX 760 outperforms the younger GeForce by about 6% and is 10% inferior to the older GTX 670. ASUS is nominally only 3% weaker than the Radeon HD 7970, and when overclocked it comes close to the heels of the Radeon GHz Edition.

The situation is changing noticeably. Now, even with overclocking, the GeForce GTX 760 is not able to catch up with a simple Radeon HD 7970. There is a 9% difference between the GeForce GTX 660 Ti and the GeForce GTX 760. The latter is inferior to the older GeForce GTX 670 by 10%. ASUS beats the reference by less than 5%.

conclusions

Based on the results of our updated list of tests, it is clear that the new GeForce GTX 760 occupies a middle position between the GeForce GTX 660 Ti and GeForce GTX 670, easily outperforming the older card when its operating frequencies are increased. At the same time, overclocking helps to reach the level of a simple Radeon HD 7970 or even slightly overtake this competitor. There are, of course, games in which GeForce cannot keep up with its AMD competitor even when overclocked. But there are also applications in which the GeForce GTX 760 is initially not inferior to its opponent or even demonstrates slight superiority.

The ASUS GTX760-DC2OC-2GD5 video card is an interesting and original solution. The cooling system provides temperatures lower than those of many top-end GeForce GTX 760 from other manufacturers, but ASUS operating frequencies are also lower. Factory overclocking is very modest. The noise level is low. The card's acoustic parameters are more pleasant than the Palit GTX 760 JetStream due to its softer sound, although the fan speed is higher. In terms of overclocking, the reviewed adapter also outperformed Palit. It is noteworthy that JetStream uses a reference design PCB. So, for all its simplicity and refusal to increase power phases, the original GeForce GTX 760 from ASUS still have certain advantages over standard design cards. But they cannot compete with the MSI N760 TF 2GD5/OC and the Inno3D iChill GTX 760 HerculeZ 3000. Realizing the full overclocking potential while maintaining acoustic comfort will be problematic. But it won’t be difficult to overclock the card to the performance level of the GeForce GTX 670 without damaging your hearing.

Testing equipment was provided by the following companies:

ASUS - ASUS PB278Q monitor, ASUSRampageIVFormula/Battlefield 3 board, GTX760-DC2OC-2GD5 and GTX670-DCMOC-2GD5 video cards;
Intel – Intel Core i7-3930K processor and Intel SSD 520 Series 240GB;
Kingston - Kingston KHX2133C11D3K4/16GX memory kit;
Syntex - Seasonic SS-750KM power supplies;
Thermalright - Thermalright Venomous X.

GeForce GTX 760 video cards were tested with the following configuration:

Motherboard: Sapphire Pure Black 990FX (BIOS v 1APEF032);
Processor: AMD FX 8-Core Black Edition FX-8320 4000 MHz (200x20, 1.425 V) or 1600 MHz when measuring the noise of the cooling system;
CPU cooling system: Cooler Master Hyper 212 Plus (CM-Long Life LED Fan 120 mm, 1000 rpm);
RAM: Samsung PC3-10600 (DDR3-1333, 7-7-7-20, 2x2 GB, dual-channel mode);
Hard drive: OCZ Vertex 4, 128 GB;
Power supply: ETG Premium 1200 Watt;
Housing: open stand.

Software

Operating system: Windows 7 x64 Ultimate;
Video card driver: NVIDIA ForceWare WHQL 320.49;
Helper utilities:

ASUS GPU-Tweak v. 2.4.3.1;
GPU-z v. 0.7.2;
Heaven Benchmark 4.0;
MSI Kombustor 2.5.

Testing tools and methodology

To overclock graphics accelerators, as well as monitor temperatures and fan speeds, the ASUS GPU-Tweak v. utility was used. 2.4.3.1.

The stability of video cards during overclocking was checked using the MSI Kombustor 2.5 utility (GPU Burn-in mode, 1280 x 1024). The obtained frequencies were additionally verified by running the Heaven Benchmark v 4.0 test with an extreme level of tessellation and graphics tests from the 3DMark 2013, 3DMark 11 and 3DMark Vantage packages.

To check the temperature conditions of video cards in conditions close to everyday conditions, Heaven BenchMark v was used. 4.0 (quality: ultra, tessellation: extreme, AA8x, 1280 x 1024).

The noise level was measured using a TM-102 digital sound level meter with a measurement error of no more than 0.5 dB. The measurements were taken from a distance of 0.5 m. The background noise level in the room was no more than 28 dB. The air temperature in the room was 25 degrees Celsius.

Standard frequencies and overclocking

The ASUS GTX 760 DirectCU II OC video card received a basic overclocking of the graphics core from 980 to 1006 MHz (including the use of GPU Boost technology - 1072 MHz), video memory operates at a frequency of 1502 MHz; under load, the accelerator can withstand a voltage of 1,200 V throughout the entire testing phase.

In idle mode, the frequencies are reset to 135 MHz/162 MHz, and the software monitoring voltage is set to 0.862 V.

Using the software, we were able to find the optimal values for the graphics core and memory, at which complete stability was observed in long-term tests - 1117 MHz (Boost up to 1182 MHz) for the graphics core and 1750 MHz for memory, respectively. Under load, the actual frequency is 1228 MHz! Regarding overclocking, this is an average result, the reason for which could be either the four-phase power supply of the GPU, or a “successful” specimen that ended up in our laboratory.

The Gigabyte GTX 760 WindForce 3X OC model also comes with a factory overclock to 1085 MHz core (in Boost mode - up to 1150 MHz). Here you can feel the confident hand of the engineers, who set the base value higher than that achieved by the reference card: let me remind you that the Inno3D GTX 760 Reference had stable operation only at a frequency of 1080 MHz, which is slightly lower than the standard frequencies of the Gigabyte video adapter! Moreover, even the voltage is lower than the reference one - 1.187 V! There were no hints of “overclocking” or lack of voltage on the GPU when the video card was running throughout the entire day of testing.

Without load, the accelerator reduces the values to the same frequency formula as ASUS - 135 MHz/162 MHz.

As you remember, the video card is equipped with an eight-phase power supply system, and this had a beneficial effect on the overclocking potential: the base frequency was raised by 100 MHz, which provided a value of 1185 MHz (Boost - 1250 MHz). And in gaming applications, where temperature and power consumption values are far from maximum, the graphics core frequency was recorded at 1306 MHz. This is simply an excellent result, which was approached only by another participant in previous testing - the GTX 760 HerculeZ 2000 4 GB OC.

The video memory clocked at 1835 MHz, which is one of the highest values among the models reviewed.

Then it was the turn of the MSI GTX 760 Twin Frozr Gaming OC. Even in standard operating mode, the graphics core frequency was increased relative to NVIDIA recommendations and amounted to 1020 MHz, while, like ASUS and Gigabyte, the manufacturer did not increase the standard value of video memory. The accelerator operates at a voltage of 1,200 V, which remains this way even under prolonged loads.

Manual overclocking managed to achieve stable operation at values of 1130 MHz (Boost - 1195 MHz) and 1800 MHz for the core/memory domains, respectively.

The MSI GTX 760 Twin Frozr Gaming OC video card has a distinctive feature that is not shared by other participants - this is the Power Target value, which the user can change up to 145%! For other accelerators it is in the range of 110-115%, but MSI is ahead of everyone. This will be useful for fans of extreme overclocking with increasing voltage above the permissible level: they will not have to modify the BIOS to increase the maximum power consumption value.

Study of the potential of the cooling system

As with the first review, testing was carried out in two stages:

With standard thermal interface;
With its replacement with Gelid GC-Extreme paste and slight overclocking.

First, measurements were taken in idle mode, load mode, and after overclocking with automatic fan control mode. The load was carried out in three runs in Heaven Benchmark 4.0, then a fixed maximum temperature and fan speed were entered into the diagram; noise values are shown in a separate graph.

Idle temperature.

Idle temperature

Standard settings
°C

Without load, the graphics accelerators reduce the graphics core frequency to 135 MHz at a voltage of 0.862 V in all three cases.

Rotation speed of fan impellers (in all cases, percentages of the maximum according to software monitoring data are indicated in parentheses).

Fan speed

Standard settings
rpm