Two video cards on one PC: how it works, why they are needed and how to connect them


Anyone who has wanted to get more performance out of their computer has undoubtedly wondered, “Can I add an extra video adapter to my motherboard and get twice the performance instead of buying a more expensive and newer graphics card?” But as a rule, the answer is no, you won’t be able to get twice the power from running two GPUs. However, it is possible to improve performance with two graphics cards, but there are some limitations and certain problems.

Advantages

There are two real benefits to running multiple graphics cards. The main reason is increased gaming performance. By having two or more graphics cards that share functions to render 3D images, computer games can run at higher FPS and resolution with additional filters. This additional capacity can significantly improve the graphics quality of games. Of course, many modern graphics cards can play the game at up to 1080p resolution. The real benefit is the ability to drive games at higher resolutions, such as on 4K displays, which have four times the resolution, or drive multiple monitors.

Another advantage is for people who want to upgrade later without replacing their graphics card. By purchasing a graphics card and a dual-card capable motherboard, you can add a second graphics card later to improve performance without having to remove your existing graphics card. The only problem with this plan is that the graphics card cycles approximately every 18 months, meaning a compatible card may be difficult to find unless you plan to purchase one within two years.

⇡#deceptive gigabytes

A significant problem in any configuration with multiple GPUs is the amount of memory on the video adapters. The fact is that the RAM of individual video cards does not add up into one large array. Instead, their content is duplicated. Thus, even if the combined computing power of two “trimmed” GPUs is comparable to the power of a single flagship core, the tandem has a frame buffer equal to the memory of one card.

And heavy graphics modes, you know, can be extremely demanding on memory. Take games known for their thirst for extra megabytes, which we use as benchmarks. Metro 2033 at a resolution of 2560x1440 with maximum settings takes up up to 1.5 GB of RAM, Far Cry 3 - 1.7, and Crysis 3 - almost 2 GB.

READ MORE: Which car compressor is better? How to choose tire inflation | Autoblog

Well, since it is precisely for such games and such settings that we are ready to resort to such a strange solution as two relatively budget cards in the SLI/CrossFireX configuration, video adapters with a RAM capacity of less than 2 GB are not suitable for us. In this regard, NVIDIA products are a little more fortunate, since even the GeForce GTX 650 Ti BOOST, according to official specifications, comes in a 2 GB memory version, not to mention the GTX 660. In addition, even the GTX 680 has no more memory (GTX only 770 versions with 4 GB GDDR are allowed 5).

The Radeon HD 7790 is designed as an adapter exclusively with 1 GB of memory, although AMD partners produce models with double frame buffer on their own initiative. The HD 7850 version with 1 GB of memory, on the contrary, officially ceased to exist due to the advent of the HD 7790. And most importantly, the flagship, Radeon HD 7970 GHz Edition, is equipped with 3 GB of memory - one and a half times more than the “average” ones from AMD.

In general, let’s stop guessing. Honestly, the performance of assembling two mid-range gaming video cards in SLI or CrossFireX is difficult to predict based on bare theory. We need to test, and we are ready for any surprises. But before moving on to practical work, let's discuss another aspect of our exotic project - energy consumption.

NVIDIA cards of comparable performance are more power hungry. GeForce GTX 650 Ti BOOST and GTX 660 have the same TDP - 140 W. When paired, such adapters are far superior to the GeForce GTX 680 (195 W) and even the GTX 770 (230 W).

Flaws

The big disadvantage of two video cards in one computer is the cost. With top-end graphics cards already costing $500 or more, many consumers are finding it difficult to afford a second one. While ATI and Nvidia offer cheaper cards with dual-card support, it's often better to spend the same amount of money on one card with equal or sometimes better performance than two inexpensive ones.

Another problem is that not all games benefit from multiple graphics cards. This situation has improved significantly since the first dual-card setups were introduced, but some graphics engines still don't handle dual graphics cards well. In fact, some games may show a slight decrease in performance compared to a single graphics card. In some cases, stuttering occurs, causing the video to appear choppy.

Two AMD Radeon video cards connected via DVI cable

Modern video cards are very energy intensive. Having two of them in a system can almost double the amount of wasted electricity required to run them in tandem. For example, a single high-end graphics card may require a 500W power supply to run properly. Having two of the same cards may require around 850W. Most desktop computers do not have such powerful power supplies. Please refer to your computer's power and requirements before purchasing more powerful hardware.

Performance benefits in a dual-card environment depend on other components of the computer system. Even with two top-end graphics cards, the lower-end processor can regulate the amount of data the system can provide to the graphics cards. As a result, dual graphics cards are usually only recommended in high-end systems.

Further development of GPU parallel connection

With the release of the new generation of GeForce RTX video cards, NVIDIA also announced a new interface (bridge) NVLink for connecting two GPUs to work together. They are connected using NVLink Bridge


. The data transfer speed over the new interface has increased 10 times compared to SLI. From 10 GB/s to 100 GB/s. NVLink is supported only by the GeForce RTX 2080 and RTX 2080 Ti; the 2070 and 2060 models do not have it.

Judging by my analysis of various gaming tests conducted with two latest generation video cards (2080Ti), the new interface in almost all modern games at 8k resolution (7680 × 4320 pixels) gives at least a 40% performance increase compared to running SLI mode on two 1080Ti. Below in the image I have shown the results of one of these tests.

How to install two video cards on a PC

In the future, we will have to plan our computer purchases very wisely. If the difference in energy consumption between 2 and 4 sticks of RAM is not significant, then when installing 2 or more video cards we will need to take into account their total power consumption.

The motherboard must have slots for installing two video cards and must support their joint operation. To find out if your motherboard supports dual graphics cards, go to the motherboard manufacturer's official page and check the specifications. This can also be determined by the presence of the Crossfire, SLI or NVLink logo on the box.

The power supply must also have enough power to pull all the computer hardware + 2 video cards. Now this requires about 900 watts.

When using two video cards, games become processor-dependent. The higher the resolution you use and the more load you put on the video card(s), the less this will appear. As the monitor resolution decreases and the load on the video card(s) decreases, this effect decreases until it disappears completely, since the video card becomes the bottleneck, not the central processor.

To work with two top-end video cards, you will have to fork out for the most powerful processor, otherwise all this is pointless.

High-quality cooling of all your system unit components must be organized, since heat generation is very high.

Link NVLink only to video cards running on the same type of GPU. The manufacturer is not important, but the frequency of the GPU is important, since when combined, both video cards will operate at a lower frequency. For example, we have one card that has an operating frequency of 1695 MHz, and the other 1770 MHz, then in joint mode they will both operate at 1695 MHz. Damn, everything is exactly like RAM.

A pair of graphics cards requires a computer case that is large enough to accommodate them and additional equipment.

How to connect two video cards to one PC

If you have built a very powerful gaming or work system and want to make it even more powerful, then purchasing a second video card will help with this. In addition, two models from the middle price segment can work better and faster than one top-end model, while costing several times less. But in order to do this, you need to pay attention to several points. Let's take a closer look at them.

What you need to know before connecting two GPUs to one PC

If you are just going to purchase a second graphics adapter and do not yet know all the nuances that need to be observed, then we will describe them in detail. Thus, during assembly you will not encounter various problems and component breakdowns.

  1. Make sure your power supply has enough power. If the video card manufacturer’s website says that it requires 150 Watts, then two models will require 300 Watts. We recommend taking a power supply with a power reserve. For example, if you currently have a 600-watt unit, but for the cards to function you need 750, then do not skimp on this purchase and buy a 1-kilowatt unit, so you will be sure that everything will work correctly even at maximum loads.

Why is this necessary?

There may be several options here.

  • first - you want to organize a workspace from several monitors and you simply do not have enough video outputs on one board;
  • Another option is to combine video cards together for the sole purpose of increasing performance.


Reasons to install 2 video cards

Well, let's consider both options.

What is needed for this?

  1. First, we need a motherboard with several slots for installing video cards. It looks like this.

Important! Be sure to check in the documentation or on the website of the board manufacturer which standard for connecting a pair of video cards it supports. If you are going to install two nVidia video cards, then you need a standard called SLI, if AMD Radeon, then the CrossFire standard.

Note! You only need to buy identical video cards and a connecting cable if you are going to combine them to get more power. If your goal is to connect more monitors, then the video cards may not match, and the cable is not needed at all.

Different GPUs on one computer

If video cards from the same manufacturer are very different from each other or they are from different manufacturers (Nvidia and AMD), then it will not be possible to combine them . But you can use them to distribute tasks. For example, you can play on one card and stream or do video editing on another.

In the case of cards from one manufacturer, it is enough to install the latest drivers and the system will work. Each board will operate discretely and perform its assigned task.

If the cards are from different companies, then the motherboard must be able to work with them simultaneously. First you need to insert the AMD card and install the drivers. After that, turn off the computer and insert the Nvidia GPU. After turning on the system, install drivers for Nvidia. Reboot, the cards will work separately.

When creating a system of two cards, you need to consider the purpose of the combination and possible costs. It is often more profitable to buy one top-end GPU than to build a cascade of two low-power ones.

Last updated October 12, 2022 at 12:46 pm

Two video cards installed on one computer and combined using SLI (Scalable Link Interface) from NVIDIA or CrossFire from AMD give a performance increase of 15-30%. Very rarely, depending on game optimization, the increase can reach 50%. AMD CrossFire technology is noticeably less optimized than NVIDIA's SLI.

Connecting two video cards for multiple monitors

The easy way. In this case, you should not count on a strong increase in productivity.

  1. Turn off your computer and unplug it. Wait about 20 seconds until the capacitors on the motherboard are completely discharged. Now you can get started.
  2. Remove the side cover of the system unit, located on the side opposite the motherboard connectors. To do this, unscrew the two fixing screws and pull it back.

Note!

Since we have opened the computer, be sure to evaluate its condition in terms of dustiness. If there is a lot of dust, carefully remove it with a vacuum cleaner or compressor. Do not scratch the boards, this may lead to their failure!

Note! In this case, you don’t even have to limit yourself to two video cards. If the motherboard has three connectors, and you need to connect more than six monitors, for example, then install three cards.

Note! If you have installed adapters of different power, then you should remember that on a monitor connected to a weaker adapter, performance will be lower than on a monitor connected to a strong adapter. In practice, the difference is almost imperceptible, however, when using demanding applications it can still be detected.

Connecting an SLI/CrossFire pair to increase performance

This method implies an increase in performance, but requires special video cards that support this function. These can be visually identified by the connector for the connecting cable at the top end.

  • repeat the steps in the first step, installing two identical video cards in adjacent slots;
  • connect the cards with a special cable, which must be purchased separately, as shown in the figure;

But now the matter does not end there. In order for our video cards to show the best performance, it is necessary to configure their interaction parameters correctly. This can be done using proprietary utilities provided by the manufacturer - as a rule, such utilities are installed on your computer along with the video card drivers.

Let's look at setting up such a configuration using the example of two nVidia video cards operating in SLI.

    Go to the “NVIDIA Control Panel” by calling the desktop context menu and selecting the appropriate item.

Ready! You have configured two video cards on your computer. You can start using it.

Video - How to enable SLI and CrossFire

Did you like the article? Save so you don't lose!

A little history

The year 1998 can be considered the beginning of the SLI era. Then the 3dfx company for the first time managed to realize the possibility of combining several video cards to solve one problem. However, the development of the AGP interface slowed down developments in this direction, since all motherboards of that time were equipped with only one slot for a video card.

SLI from 3DFX - the “grandfather” of modern technology

After 3dfx was purchased by Nvidia in 2001, all developments in this direction were waiting in the wings, which came only in 2004. It was then, in the 6000 series GeForce (the first video cards designed specifically for the then latest PCI-Express interface) that the existing developments were improved and introduced into consumer products.

SLI from two GeForce 6600GT

At first, it was possible to combine two GeForce 6600 or 6800 video cards, with the release of the 7000 series (which introduced the first full-fledged dual-processor board from Nvidia, 7900GX2), it became possible to build a four-processor configuration (from 2 cards), and later it became possible to combine 3 or 4 separate GPUs.

Combination technologies and algorithms for their functioning

Two GPU manufacturers: Nvidia and AMD use similar principles when connecting two video adapters, but there are differences.

SLI from Nvidia

The abbreviation "SLI" was first used by 3dfx in 1998 . It stood for Scan Line Interleave , which translated means “alternating lines.” 3dfx wanted to combine 2 graphics chips to achieve the then huge resolution of 1024 x 768. Each chip processed an even or odd line, and then all the lines were added together. But shortcomings and limitations of the AGP graphics port did not allow the technology to be successfully implemented. In 2001, 3dfx was purchased by Nvidia.

After the development of PCI-Express, it became possible to connect 2 or more video cards in one computer. Nvidia revives the project in 2004 under the same abbreviation, but the operating principle is different. SLI now stands for Scalable Link Interface , which is translated as “scalable interface”. In this technology, the process of interaction between maps and graphics generation occurs according to the following algorithms:

  1. Split Frame Rendering ( SFR ) or Separate Frame Visualization . A simple method, its essence is that each image frame is divided equally between GPUs. Each part is processed separately. Then they add up, as a result the overall fps increases. The disadvantage of this rendering is that in the real “picture” its effectiveness will decrease. The first GP may get a simple area, without a large number of details and effects, and for it the processing time will be minimal. For the second, on the contrary, part of the frame may be replete with small details; it will require more time for visualization. Because of this, the first card will be “idle” for a certain time, and the overall fps will be lower.
  1. Alternate Frame Rendering (AFR) or Alternating Frame Rendering . To avoid rendering lag issues, AFR mode is used. Here, even image maps are processed by one video processor, odd ones by another. Frames can be of varying complexity, so there is also a drop in fps. This method requires more video memory to remember previous and subsequent frames.
    There are large delays between frames, so this algorithm does not find support among users. CrossFire from AMD The development of Crossfire was started by ATIRadeon in 1999. Initially, the mode of combining two video processors was called MAXX. In 2004 the name changed to Crossfire. After the purchase of ATI Radeon by AMD, the technology acquired the name AMD CrossFireX. AMD also has several GPU operating algorithms, some similar to Nvidia modes:
  2. Scissor ( Slicing) or Slicing . The algorithm is similar to SFR for Nv />
  3. AFR . Similar to Nvidia, right down to the name.

  4. SuperAA (AntiAliasing) or Super Anti-Aliasing . The mode is similar to SLIAntiAliasing, but it uses SSAA anti-aliasing, whereas SLI uses MSAA. The AMD version gives a slightly better image, but loads the card more. Also, CrossFire has a maximum smoothing coefficient of x14, while SLI has a maximum of 32.

How to combine video cards in SLI: requirements

As with CrossFire, it is not enough to buy two Nvidia video cards to configure them in SLI. There are a number of computer requirements that must be met to ensure the normal functioning of the bundle.

  • Compatible video cards
    . As with CrossFire, the boards do not have to be identical, but built on the same processor. That is, you can combine GeForce GTX960 video cards from Asus and Gigabyte, but you cannot build SLI from GTX750Ti and GTX960 from Palit. It is recommended to use GPUs with identical memory sizes. A configuration of 1 and 2 GB cards, even if it works, will use less volume, and with differences in memory speed (for example, 2 GB DDR3 and 1 GB GDDR5), there is a risk of software errors leading to incorrect image formation. To ensure SLI functionality, cards equipped with an SLI interface and connected using a bridge are required. It is also possible to combine them programmatically, but in this case the system performance will be lower.
  • Suitable motherboard
    . In terms of chipset support, the situation with SLI is better than with CrossFire. Advanced Micro Devices decided to do without a “speak in the wheels” for gamers using boards based on competitors’ chipsets, so AMD 900 series chipsets support both technologies for combining video cards. Intel chips are also compatible with SLI, which is important. Of course, Nvidia couldn’t ignore the “native” NForce series chipsets, the top versions of which are made SLI-compatible from the start. Of course, having a suitable chipset is not enough: the board itself must have the appropriate number of PCI-Express slots. It is recommended to use products designed for gamers, where each slot has 16 or 8 PCI-E lanes. In the case of budget motherboards, you should be especially careful: the vast majority of gamers who buy such gadgets never use SLI functions, and they buy motherboards with two PCI-E slots “just in case.” Manufacturers know this and therefore often make the second connector inferior: they connect 4, 2 or even 1 line to it, instead of 16 or 8. SLI on such boards, even if it functions (not a fact: you need to look at the SLI logo on the board box), is possible video cards will not be fully opened.

Despite the presence of three full-size X16 slots, all 16 lines are connected only to the first one. Slot 2 is equipped with eight lines, slot 3 - only four

Which video cards can be combined into SLI

Connecting two video cards with an SLI bridge is possible only if an interface is provided for this bridge on the board. You can also try to combine them using software, but in this case you cannot do without “dancing with a tambourine.” And it makes no sense to build a tandem of two GT610 or GT720. One card, the speed of which will be 2 or 3 times higher than that of a similar bundle, will cost less than a set of a pair of office cards and a compatible motherboard. That is, the lack of an SLI interface on the younger representatives of graphics accelerators is not caused by the greed of the manufacturer. Nobody puts bridges, because it will make the map more expensive, but there will be no point in them. Thus, connecting two SL I video cards is possible for mid- and top-class models. These include GPUs whose second digit is 5 or higher (GTX5 5

0Ti, GTX9
6
0, GTX6
7
0, GTX7
8
0, etc.).

GeForce GT 720 does not have a bridge for SLI

Before you connect two SLI video cards, you should make sure that they are built on the same version of the graphics processor. For example, GeForce GTX650 and GTX650Ti, despite the similarity of names, operate on the basis of completely different GPUs and therefore cannot work in tandem.

⇡#do we need it?

But let’s try to find some practical benefit in such an experiment. In theory, a tandem of weak video cards can be a better purchase than one top-end adapter. The older AMD models are now the permanent Radeon HD 7970 and its overclocked brother HD 7970 GHz Edition. Among NVIDIA products, a similar pair can be considered the GeForce GTX 680 and GTX 770. Well, the Radeon HD 7790 and HD 7850 (for AMD), GTX 650 Ti BOOST and GTX 660 (for NVIDIA) can be considered as their approximate halves.

READ MORE: How to choose and use decking oil, expert advice

Let's compare the retail prices of these models. Radeon HD 7790 with a standard memory capacity of 1 GB can be purchased in Moscow online stores for an amount of 4,100 rubles, HD 7850 2 GB - for 4,800 rubles. At the same time, the Radeon HD 7970 is sold at a price of 9,800 rubles, which is only 200 rubles more expensive than a pair of HD 7850. But the HD 7970 GHz Edition is already difficult to find for less than 14,000 rubles.

Thus, CrossFireX on two inexpensive AMD cards seems a dubious idea, if only because the Radeon HD 7850, although it is half the HD 7970 in terms of the number of active computing units in the GPU, is significantly inferior to the latter in terms of core and memory clock speeds. The HD 7790 runs the processor at 1 GHz, which partly compensates for its more modest configuration compared to the HD 7850, but the weak point of the HD 7790 is the 128-bit memory bus, and most importantly, its 1 GB capacity, which will certainly affect performance in heavy environments. graphic modes (and it is for them that we are assembling such a tandem).

From the point of view of a pure experiment, not burdened by financial considerations, in terms of the configuration of computing units and GPU frequency, the combined power of a pair of low-end AMD adapters capable of operating in CrossFireX mode (we don’t take into account the Radeon HD 7770; an assembly of two of these would be completely ridiculous) , is a little short of the flagship mark.

ModelRadeon HD 7790Radeon HD 7850HD 7970HD 7970 GHz Edition
Main components
GPUBonairePitcairn ProTahiti XTTahiti XT
Number of transistors, million2080280043134313
Technical process, nm28282828
GPU clock frequency, MHz: Base Clock / Boost Clock1000/ND860/ND925/ND1000/1050
Stream processors896102420482048
Texture blocks5664128128
ROP16323232
Video memory: type, volume, MBGDDR5, 1024GDDR5, 2048GDDR5, 3072GDDR5, 3072
Memory bus width, bits128256384384
Memory clock frequency: real (effective), MHz1500 (6000)1200 (4800)1375 (5500)1500 (6000)
InterfacePCI-Express 3.0 x16
Image output
Interfaces1 x DL DVI-I, 1 x DL DVI-D, 1 x HDMI 1.4a, 1 x DisplayPort 1.21 x DL DVI-I, 1 x HDMI 1.4a, 2 x Mini DisplayPort 1.2
Max. permission VGA: 2048x1536, DVI: 2560x1600, HDMI: 4096x2160, DP: 4096x2160
Typical power consumption, W85130250250
Average retail price, rub.4 1004 8009 80014 000

The price range among NVIDIA video cards is more favorable for assembling a tandem of two low-performance SKUs compared to a single powerful adapter. GeForce GTX 650 Ti BOOST can be purchased at a price starting from 5,200 rubles, GeForce GTX 660 - for 5,800 rubles. Both have two gigabytes of onboard memory.

At the same time, prices for the GeForce GTX 680 and the newly arrived GTX 770, which is actually an updated version of the GTX 680, start at 13,000 rubles. According to the specifications, the GTX 650 Ti BOOST, the youngest graphics accelerator in NVIDIA's range that supports SLI, corresponds to half of the GTX 680 (GTX 770) and even has a larger number of ROPs, but only operates at lower frequencies.

ModelGeForce GTX 650 Ti BOOSTGeForce GTX 660GTX 680GTX 770
Main components
GPUGK106GK106GK104GK104
Number of transistors, million2 5402 5403 5403 540
Technical process, nm28282828
GPU clock frequency, MHz: Base Clock / Boost Clock980/1033980/10331006/10581046/1085
Stream processors76896015361536
Texture blocks6480128128
ROP24243232
Video memory: type, volume, MBGDDR5, 1024/2048GDDR5, 2048GDDR5, 2048GDDR5, 2048
Memory bus width, bits1502 (6008)1502 (6008)1502 (6008)1753 (7010)
Memory clock frequency: real (effective), MHz192192256256
InterfacePCI-Express 3.0 x16
Image output
Interfaces1 x DL DVI-I, 1 x DL DVI-D, 1 x HDMI 1.4a, 1 x DisplayPort 1.2
Max. permission
Typical power consumption, W140140195230
Average retail price, rub.5 2005 80013 00013 000

Is there any practical benefit to SLI?

Before combining video cards in SLI, it is advisable to familiarize yourself with the experience of using such configurations by other gamers and study the specifications of your cards, as well as the characteristics of more productive solutions in the line. Often, the two-card configuration does not provide any benefit either in terms of economy or performance. For example, two GTX950s in SLI demonstrate results comparable to one GTX970. The difference in price (about $200 and $400, respectively) of the cards themselves is justified, but if you take into account the additional costs of a powerful power supply, a dual-slot motherboard, and a high-quality and well-ventilated case, it looks dubious.

The situation is completely different if connecting two SLI video cards is part of the procedure for upgrading an existing PC purchased a couple of years ago. Cards of the GTX650Ti or GTX750 level can still be found on sale at a reasonable price (the difference with the new models is approximately equivalent to the differences in performance), and technical progress over 3 years has not brought anything revolutionary for video cards (except for HBM memory, presented by AMD, but Nvidia is not yet concerned). Therefore, adding another GPU is a completely rational step for such players.

⇡#test participants, details

Most of the cards in multiprocessor assemblies are provided for testing by NVIDIA partners, and differ from the reference design adapters in clock speeds. For the purity of the study, they had to be brought to reference specifications. However, showing respect for the work of manufacturers, below we will briefly list the advantages of each of the testing participants.

  • AMD Radeon HD 7850
  • ASUS GeForce GTX 650 Ti BOOST DirectCU II OC (GTX650TIB-DC2OC-2GD5)
  • ASUS Radeon HD 7790 DirectCU II OC (HD7790-DC2OC-1GD5)
  • Gainward GeForce GTX 650 Ti BOOST 2GB Golden Sample (426018336-2876)
  • GIGABYTE Radeon HD 7790 (GV-R779OC-1GD)
  • MSI GeForce GTX 660 (N660 TF 2GD5/OC)
  • ZOTAC GeForce GTX 660 (ZT-60901-10M)

ASUS GeForce GTX 650 Ti BOOST DirectCU II OC (GTX650TIB-DC2OC-2GD5)

ASUS sent us a slightly overclocked version of the GeForce GTX 650 Ti BOOST 2GB with the base core clock increased by 40 MHz compared to the reference specification. The board is equipped with the well-known DirectCU II cooling system, which includes heat pipes that are in direct contact with the GPU crystal. Hence the name DirectCU. But, unlike more expensive models, the GTX 650 Ti BOOST cooler has only two heat pipes, not three.

Gainward GeForce GTX 650 Ti BOOST 2GB Golden Sample (426018336-2876)

Gainward also produces an overclocked version of the GeForce GTX 650 Ti BOOST, with the difference from the ASUS implementation that the base GPU frequency is increased by only 26 MHz, but there is an increase of 100 MHz to the resulting video memory frequency. The video card cooler is also an open type and contains two copper heat pipes.

MSI GeForce GTX 660 (N660 TF 2GD5/OC)

The GeForce GTX 660 by MSI is equipped with a solid Twin Frozr III cooling system with three heat pipes. The manufacturer increased the base clock frequency of the GPU by 53 MHz relative to the reference value, without affecting the video memory frequency.

ZOTAC GeForce GTX 660 (ZT-60901-10M)

This video card has already appeared on 3DNews in the initial testing of the GeForce GTX 660. There is a slight overclocking of the GPU: from a base frequency of 980 to 993 MHz. The cooling system is an open type design with two heat pipes and a copper plate at the base.

AMD Radeon HD 7850

This is a one hundred percent reference video adapter provided by AMD itself. Accordingly, the core and memory frequencies correspond to the reference ones with an accuracy of one megahertz. Unlike all other test participants, the reference Radeon HD 7850 is equipped with a turbine cooler. The radiator has a copper base and is penetrated by three heat pipes.

ASUS Radeon HD 7790 DirectCU II OC (HD7790-DC2OC-1GD5)

The overclocker version of the Radeon HD 7790 from ASUS is overclocked by 75 MHz in processor frequency and 400 MHz in video memory. In general, AMD GPUs allow for greater factory overclocking compared to their NVIDIA counterparts. The card's cooling system is similar to that of the above-described ASUS GeForce GTX 650 Ti BOOST DirectCU II OC. This is a simplified version of the DirectCU II cooler with two heat pipes pressed against the GPU die.

GIGABYTE Radeon HD 7790 (GV-R779OC-1GD)

This card is similar in configuration to the Radeon HD 7990 from ASUS: the GPU operates at the same frequency of 1075 MHz, but the memory frequency is left at the standard 6000 MHz. The cooler is a simple design consisting of a solid aluminum radiator and a large fan. The GIGABYTE adapter cannot boast of heat pipes or an additional metal radiator base.

READ MORE: The best fitness centers in Kazan: addresses, phone numbers, reviews and prices

SLI: connecting video cards

If the computer meets the requirements for creating an SLI configuration, a suitable second video card has been purchased, you can proceed to assembly. After turning off the PC, you should remove the cover of the system unit, install the second card in the appropriate slot, connect an additional power cable to it (if required) and connect the two cards with the bridge included in the kit. The hardware connection of two video cards in SLI is now complete and you can move on to software configuration.

Power cables connected to SLI link

Connecting two SLI video cards in Windows

In order for a combination of two GPUs to function normally, you need to install Windows version of Vista or later (7, 8, 8.1 or 10) on your PC. It is also recommended to download the latest version of the driver for your video card from the Nvidia website. After this, you can proceed to configuration.

  • You need to right-click on the desktop and select “Nvidia Control Panel” in the window that appears.
  • In the “3D Settings” submenu, select “Configure SLI, PhysX”.
  • In the tab that opens, you need to check the “Maximum 3D performance” checkbox.

Enabling SLI on Windows

After applying the settings, the bundle is ready for use. It is possible that in order to correctly configure SLI operation in some games, you will have to set special parameters for them in the “Program Settings” submenu. But, as a rule, most modern software does not need this.

A big fan of high-quality Chinese technology, a lover of clear screens. Supporter of healthy competition between producers. He closely follows news in the world of smartphones, processors, video cards and other hardware.

Rating
( 2 ratings, average 5 out of 5 )
Did you like the article? Share with friends:
For any suggestions regarding the site: [email protected]
Для любых предложений по сайту: [email protected]