When it comes to any interfaces in the context of computer systems, you need to be very careful so as not to "run into" incompatible interfaces for the same components within the system.
Fortunately, when it comes to the PCI-Express interface for connecting a video card, there will be practically no problems with incompatibility. In this article, we will analyze this in more detail, and also talk about what this PCI-Express is.
What is PCI-Express for and what is it?
Let's start, as usual, from the very beginning. PCI-Express (PCI-E) interface Is a means of interaction, in this context, consisting of a bus controller and a corresponding slot (Fig. 2) on motherboard (to summarize).
This high-performance protocol is used, as noted above, to connect a video card to the system. Accordingly, there is a corresponding PCI-Express slot on the motherboard, where the video adapter is installed. Previously, video cards were connected via the AGP interface, but when this interface, to put it simply, “was no longer enough,” PCI-E came to the rescue, the detailed characteristics of which we will now talk about.
Figure 2 (PCI-Express 3.0 slots on the motherboard)
Main characteristics of PCI – Express (1.0, 2.0 and 3.0)
Despite the fact that the names PCI and PCI-Express are very similar, the principles of connection (interaction) are fundamentally different. In the case of PCI-Express, a line is used - a bi-directional serial connection, point-to-point, these lines can be several. In the case of video cards and motherboards (we do not take into account Cross Fire and SLI) that support PCI-Express x16 (that is, the majority), you can easily guess that there are 16 such lines (Fig. 3), quite often on motherboards with PCI- E 1.0, you could see a second x8 slot for SLI or Cross Fire.
Well, in PCI, the device connects to a common 32-bit parallel bus.
Figure: 3. An example of slots with different numbers of lines
(as mentioned earlier, x16 is the most commonly used)
For the interface, the bandwidth is 2.5 Gbps. We need this data to track changes in this parameter in different PCI-E versions.
Further, version 1.0 evolved into PCI-E 2.0... As a result of this transformation, we got twice the bandwidth, that is, 5 Gb / s, but I would like to note that the graphics adapters did not particularly benefit in performance, since this is just a version of the interface. Most of the performance depends on the video card itself, the version of the interface can only slightly improve or slow down the data transfer (in this case, there is no "braking", and there is a good margin).
In the same way, in 2010, with a margin, the interface was developed PCI-E 3.0, at the moment it is used in all new systems, but if you still have 1.0 or 2.0, then do not worry - below we will talk about the relative backward compatibility of different versions.
In PCI-E 3.0, the bandwidth has been doubled compared to 2.0. There were also many technical changes made there.
Expected to be born by 2015 PCI-E 4.0, which is absolutely not surprising for a dynamic IT industry.
Well, okay, let's finish with these versions and bandwidth numbers, and touch on a very important issue of backward compatibility of various PCI-Express versions.
PCI-Express 1.0, 2.0, and 3.0 backward compatibility
This question worries many, especially when choosing a video card for the current system. Since being content with a system with a motherboard that supports PCI-Express 1.0, there are doubts whether a video card with PCI-Express 2.0 or 3.0 will work correctly? Yes, it will, at least so the developers who have provided this very compatibility promise. The only thing is that the video card will not be able to fully reveal itself in all its glory, but the performance losses, in most cases, will be insignificant.
Quite the opposite, you can safely install video cards with the PCI-E 1.0 interface in motherboards that support PCI-E 3.0 or 2.0, nothing is limited here at all, so be sure about compatibility. If, of course, everything is in order with other factors, these include an insufficiently powerful power supply unit, etc.
In general, we have talked in some detail about PCI-Express, which will allow you to get rid of a lot of ambiguities and doubts about compatibility and understanding of differences in PCI-E versions.
I have been asked this question more than once, so now I will try to give the most accessible and brief answer to it, for this I will give pictures of PCI Express and PCI expansion slots on the motherboard for a clearer understanding and, of course, I will indicate the main differences in characteristics, i.e. .e. very soon, you will find out what these interfaces are and how they look.
So, first, let's briefly answer this question, what exactly are PCI Express and PCI.
What are PCI Express and PCI?
PCI Is a computer parallel I / O bus for connecting peripheral devices to the computer's motherboard. PCI is used to connect: video cards, sound cards, network cards, TV tuners and other devices. The PCI interface is outdated, so you probably won't be able to find, for example, a modern video card that connects via PCI.
PCI Express (PCIe or PCI-E) is a computer serial I / O bus for connecting peripherals to the computer's motherboard. Those. at the same time, a bi-directional serial connection is already used, which can have several lines (x1, x2, x4, x8, x12, x16 and x32) the more such lines, the higher the bandwidth of the PCI-E bus. The PCI Express interface is used to connect devices such as: video cards, sound cards, network cards, SSD drives and others.
There are several versions of the PCI-E interface: 1.0, 2.0 and 3.0 (version 4.0 is coming soon). This interface is usually designated, for example, like this PCI-E 3.0 x16which means PCI Express 3.0 version with 16 lanes.
If we talk about whether, for example, a video card that has a PCI-E 3.0 interface on a motherboard that only supports PCI-E 2.0 or 1.0 will work, so the developers say that everything will work, but of course, take into account that the bandwidth will be limited by the capabilities of the motherboard. Therefore, in this case, I think it's not worth overpaying for a video card with a newer version of PCI Express ( if only for the future, i.e. You are planning to purchase a new motherboard with PCI-E 3.0). Also, and vice versa, suppose your motherboard supports PCI Express 3.0, and your video card supports, say, 1.0, then this configuration should also work, but only with PCI-E 1.0 capabilities, i.e. there is no limitation here, since the video card in this case will work at the limit of its capabilities.
Differences PCI Express from PCI
The main difference in characteristics is, of course, the bandwidth, for PCI Express it is much higher, for example, PCI at 66 MHz has a bandwidth of 266 Mb / s, and PCI-E 3.0 (x16) 32 GB / sec.
Externally, the interfaces are also different, so you won't be able to connect, for example, a PCI Express video card to a PCI expansion slot. PCI Express interfaces with different number of lanes also differ, I'll show all this in the pictures.
PCI Express and PCI expansion slots on motherboards
PCI and AGP slots
PCI-E x1, PCI-E x16 and PCI slots
PCI Express interfaces on video cards
That's all for me, bye!
- Hello! Please explain the difference in bandwidth between PCI Express 3.0 x16 and PCI Express 2.0 x16. Now there are still on sale motherboards with the PCI Express 2.0 x16 interface. I am with i will greatly lose in performance if I install a new video interface cardPCI Express 3.0 to a computer with a motherboard with only a slotPCI-E 2.0? I think that I will lose, because the totalbaud ratepCI Express 2.0 equals - 16 GB / s, and the totalthe data transfer speed of PCI Express 3.0 is twice as high -32 GB / s
- Hello! I have a computer with a powerful, but not new, Intel Core i7 2700K processor and a motherboard that has a PCI Express 2.0 slot. Tell me, if I buy a new PCI Express 3.0 video card, this video card will work twice as slow as if I had a motherboard with a connector PCI Express 3.0? So it's time for me to change my computer?
- Please answer this question. My motherboard has two connectors: PCI Express 3.0 and PCI Express 2.0, but the slot PCI Express 3.0 new graphics cardPCI Express 3.0 does not climb, the radiator of the south bridge interferes. If I install a graphics cardPCI-E 3.0 to slot PCI-E 2.0, will my video card perform worse than if it were installed in a PCI Express 3.0 slot?
- Hello, I want to buy a little used motherboard from a friend for two thousand rubles. Three years ago he bought it for 7000 rubles, but I am confused by the fact that it has a slot for an interface video card PCI-E 2.0, and I have a video cardPCI-E 3.0. Will my graphics card on this motherboard run at full capacity or not?
Hello friends! Today on sale you can find motherboards with a connector for installing PCI Express 2.0 x16 video cards, and PCI Express 3.0 x16. The same can be said about graphics adapters, there are video cards on sale with an interfacePCI-E 3.0 as well as PCI-E 2.0. If you look at the official specifications of the PCI Express 3.0 x16 and PCI Express 2.0 x16 interfaces, you will find out that the total data transfer rate for PCI Express 2.0 is - 16 GB / s, while PCI Express 3.0 is twice as large -32 GB / s I will not delve into the jungle of the specifics of these interfaces and just tell you that there is such a big difference the data transfer rate is visible only in theory, in practice it is very small.If you read articles on this topic on the Internet, thenyou will conclude that modern PCI Express 3.0 graphics cards operate at the same speed in PCI Express 3.0 x16 and PCI Express 2.0 x16 slots anddifference in throughputbetween PCI-E 3.0 x16 and PCI-E 2.0 x16 is only 1-2% loss of graphics card performance. That is, it doesn't matter which slot you install the video card into, PCI-E 3.0 or PCI-E 2.0, everything will work the same.
But unfortunately, all these articles were written in 2013 and 2014, and at that time there were no games like Far Cry Primal, Battlefield 1 and other new products that appeared in 2016. Also in 2016 was released the family of NVIDIA 10-series GPUs, for example, the GeForce GTX 1050 and GeForce GTX 1050 Ti graphics cards, and even GTX 1060. My experiments with new games and new video cards have shown that the advantage of the PCI-E 3.0 interface overPCI-E 2.0 is no longer 1-2%, but on average 6-7%. What is interesting if the video card is lower in class thanGeForce GTX 1050 , then the percentage is less (2-3%) , and if on the contrary, then more -9-13%.
So, in my experiment I used a video card GeForce GTX 1050 PCI-E 3.0 interface and motherboard with connectorsPCI Express 3.0 x16 and PCI Express 2.0 x16.
H graphics settings in games are maximum everywhere.
- FAR CRY PRIMAL game. Interface PCI-E 3.0 has shown an advantage overPCI-E 2.0 since always higher by 4-5 frames, which is about4 % %.
- Battlefield 1 game. The gap between PCI-E 3.0 and PCI-E 2.0 was8-10 frames , which is in a percentage ratio of about 9%.
- Rise of the Tomb Raider. Advantage of PCI-E 3.0 averages 9-10 fps or 9%.
- The witcher. The advantage of PCI-E 3.0 was 3%.
- Grand Theft Auto V. The advantage of PCI-E 3.0 is 5 fps or 5%.
That is, the difference in bandwidth between the PCI-E 3.0 x16 and PCI-E 2.0 x16 interface is still not in favor PCI-E 2.0. Therefore, I would not buy a motherboard with one PCI-E 2.0 slot at the moment.
A friend of mine bought a used motherboard for three thousand rubles. Yes, once it was heaped up and cost about ten thousand rubles, it has a lot of connectorsSATA III and USB 3.0, also 8 slots for RAM, it supports RAID technology, etc., but it is built on an outdated chipset and a slot for a video card on it PCI Express 2.0! My opinion would be better to buy. Why?
It may well happen that in a year or two, the latest video cards will work only in the slotPCI Express 3.0 x16 , and your motherboard will have an obsolete and already unused connectorPCI Express 2.0 x16 ... You buy a new video card, and it will refuse to work in the old connector. Personally, I have already come across many times that the video card PCI-E 3.0 did not start on the mat. board with connector PCI-E 2.0, and even updating the motherboard BIOS did not help.Also I dealt with video cardsPCI-E 2.0 x16 that refused to work on older motherboards with an interfacePCI-E 1.0 x16, although everywhere they write about backward compatibility.Cases when the PCI Express 3.0 x16 video card did not start on motherboards withPCI Express 1.0 x16, even more.
Well, do not forget about the appearance this year of the interface PCI Express 4.0. In this case, PCI Express 3.0 will be obsolete.
The PCI Express standard is one of the foundations of modern computers. PCI Express slots have long occupied a solid place on any motherboard in a desktop computer, replacing other standards such as PCI. But even the PCI Express standard has its own varieties and different connection patterns. On new motherboards, starting from about 2010, you can see on one motherboard a whole scattering of ports designated as PCIE or PCI-E, which can differ in the number of lines: one x1 or several x2, x4, x8, x12, x16 and x32.
So let's find out why there is such confusion among the seemingly simple PCI Express peripheral port. And what is the purpose of each PCI Express x2, x4, x8, x12, x16 and x32 standard?
What is PCI Express Bus?
In the distant 2000s, when the transition from the aging PCI standard (ext. - interconnection of peripheral components) to PCI Express took place, the latter had one huge advantage: instead of the serial bus, which was PCI, a point-to-point access bus was used. This meant that each individual PCI port and the cards installed in it could take full advantage of the maximum bandwidth without interfering with each other, as was the case when connected to PCI. In those days, the number of peripheral devices inserted into expansion cards was enough. Network cards, audio cards, TV tuners, and so on - all required a fair amount of PC resources. But unlike the PCI standard, which used a common bus for transferring data with several devices connected in parallel, PCI Express, when considered in general, is a packet network with a star topology.
PCI Express x16, PCI Express x1, and PCI on one card
In layman's terms, imagine your desktop PC as a small store with one, two sellers. The old PCI standard was like a grocery store: everyone waited in the same queue to be served, experiencing speed issues with a limited one seller at the counter. PCI-E is more like a hypermarket: each customer follows their own individual route for groceries, and several cashiers take the order at the checkout.
Obviously, a hypermarket is several times faster than a regular store in terms of service speed, due to the fact that the store cannot afford the bandwidth of more than one seller with one checkout.
Also with dedicated data lanes for each expansion card or built-in motherboard components.
Impact of the number of lines on throughput
Now, to expand on our store-and-hypermarket metaphor, imagine that each department of the hypermarket has its own cashiers, reserved only for them. This is where the idea of \u200b\u200bmultiple data lanes comes in.
PCI-E has gone through many changes since its inception. Currently, new motherboards usually use version 3 of the standard already, with the faster version 4 becoming more common, with version 5 expected in 2019. But different versions use the same physical connections, and these connections can be made in four basic sizes: x1, x4, x8 and x16. (x32 ports exist, but are extremely rare on motherboards of regular computers).
Different physical sizes of PCI-Express ports allow them to be clearly divided according to the number of simultaneous connections to the motherboard: the larger the port is physically, the more maximum connections it can transfer to the card or vice versa. These compounds are also called lines... One line can be thought of as a track consisting of two signal pairs: one for sending data and the other for receiving.
Different versions of the PCI-E standard allow different speeds for each lane. But generally speaking, the more lanes there are on a single PCI-E port, the faster data can flow between the peripheral and the rest of the computer.
Returning to our metaphor: if we are talking about one seller in a store, then the strip x1 will be this only seller serving one customer. The store with 4 cashiers already has 4 lines x4... And so on, you can list cashiers by the number of lines, multiplying by 2.
Various PCI Express Cards
Device types using PCI Express x2, x4, x8, x12, x16, and x32
For the PCI Express 3.0 version, the total maximum data transfer rate is 8 GT / s. In reality, the speed for the PCI-E 3 version is slightly less than one gigabyte per second per lane.
Thus, a device using a PCI-E x1 port, such as a low-power sound card or a Wi-Fi antenna, will be able to transmit data at a maximum speed of 1 Gbps.
A card that physically fits into a larger slot - x4 or x8a USB 3.0 expansion card, for example, can transfer data four or eight times faster, respectively.
The transfer rate of PCI-E x16 ports is theoretically limited by the maximum bandwidth of about 15 Gb / s. This is more than enough in 2017 for all modern graphics cards developed by NVIDIA and AMD.
Most discrete graphics cards use a PCI-E x16 slot
PCI Express 4.0 allows you to use 16 GT / s, and PCI Express 5.0 will use 32 GT / s.
But there are currently no components that can use that many lanes with maximum bandwidth. Modern top-end graphics cards usually use the x16 PCI Express 3.0 standard. It makes no sense to use the same lanes for a network card that will use only one line on the x16 port, since the Ethernet port is only able to transfer data up to one gigabit per second (which is about one eighth of the bandwidth of one PCI-E lane - remember: eight bits in one byte).
You can find PCI-E SSDs on the market that support the x4 port, but they seem to be soon supplanted by the rapidly evolving new M.2 standard. for solid-state drives that can also use the PCI-E bus. High quality NICs and enthusiast hardware such as RAID controllers use a mix of x4 and x8 formats.
PCI-E Port and Lane Sizes May Vary
This is one of the most confusing tasks for PCI-E: a port can be made in x16 form factor, but have insufficient bandwidth to pass data, for example, just for example x4. This is because even though PCI-E can carry an unlimited number of individual connections, there is still a practical limit on the chipset's bandwidth. Cheaper motherboards with more budget chipsets can only have one x8 slot, even though that slot can physically house an x16 card.
In addition, gamer-oriented motherboards include up to four full PCI-E slots with x16 and the same number of lanes for maximum bandwidth.
Obviously this can cause problems. If the motherboard has two x16 slots, but one of them has only x4 strips, then connecting a new graphics card will reduce the performance of the first by as much as 75%. This is, of course, only a theoretical result. The architecture of the motherboards is such that you will not see a dramatic drop in performance.
Correct configuration of two video graphics cards should use exactly two x16 slots if you want maximum comfort from a tandem of two video cards. To find out how many lines on your motherboard one or another slot has, the manual at the office will help. manufacturer's website.
Sometimes manufacturers even mark the number of lines on the PCB of the motherboard next to the slot
Be aware that a shorter x1 or x4 card can physically fit into a longer x8 or x16 slot. The contact configuration of the electrical contacts makes this possible. Naturally, if the card is physically larger than the slot, then it will not work to insert it.
Therefore, remember that when buying expansion cards or upgrading current ones, you must always remember both the size of the PCI Express slot and the number of lanes required.
Introduction
Moore's Law states that the number of transistors on a silicon chip that is profitable to manufacture doubles every couple of years. But don't think that processor speed doubles every couple of years too. Many people have this misconception, and users often expect PC performance to scale exponentially.
However, as you may have noticed, the top processors on the market have been stuck between 3 and 4 GHz for six years now. And the computer industry had to look for new ways to increase computing performance. The most important of these methods is to maintain a balance between platform components that use the PCI Express bus, an open standard that allows high-speed graphics cards, expansion cards, and other components to exchange information. And PCI Express is just as important to performance scaling as multi-core processors. While dual-core, quad-core and six-core processors can only be loaded with multithreading-optimized applications, any program installed on your computer interacts in one way or another with components connected via PCI Express.
Many journalists and experts expected next-generation PCI Express 3.0 motherboards and chipsets to appear in the first quarter of 2010. Unfortunately, backward compatibility issues delayed PCI Express 3.0, and today it's been half a year, but we are still waiting official information on the publication of the new standard.
However, we spoke with the PCI-SIG (Special Interest Group, which is responsible for PCI and PCI Express), which allowed us to get some answers.
PCI Express 3.0: plans
Al Yanes, President and Chairman of PCI-SIG, and Ramin Neshati, Chairman of PCI-SIG Serial Communications Workgroup, shared their current plans for PCI Express 3.0.
 |
Click on the picture to enlarge.
On June 23, 2010 version 0.71 of the PCI Express 3.0 specification was released. Yans argued that version 0.71 should fix any backward compatibility issues that caused the initial delay. Neshati noted that the main compatibility issue was the "DC wandering" feature, which he explained as PCI Express 2.0 and earlier devices "did not give the necessary zeros and ones" to match the PCI Express 3.0 interface.
Today, with backward compatibility issues resolved, PCI-SIG is poised to release baseline 0.9 "later this summer". And behind this basic version, version 1.0 is expected in the fourth quarter of this year.
Of course, the most intriguing question is when will PCI Express 3.0 motherboards hit store shelves. Neshati noted that he expects the first products to appear in the first quarter of 2011 (triangle "FYI" in the picture with the plan).
Neshati added that there should be no silicon-level changes between versions 0.9 and 1.0 (that is, all changes will only affect software and firmware), so some products should reach the market before the final 1.0 specification appears. And products can already be certified for the PCI-SIG "Integrator's List" (triangle "IL"), which is a variant of the PCI-SIG compliance logo.
Neshati jokingly named the third quarter of 2011 as the date "Fry's and Buy" (probably referring to Frys.com, Buy.com or Best Buy). That is, during this period we should expect the appearance of a large number of products with PCI Express 3.0 support in retail stores and online stores.
PCI Express 3.0: Designed for Speed
For end users, the main difference between PCI Express 2.0 and PCI Express 3.0 will be a significant increase in maximum bandwidth. PCI Express 2.0 has a signal transfer rate of 5 GT / s, that is, the bandwidth is 500 MB / s for each line. Thus, the main PCI Express 2.0 graphics slot, which typically uses 16 lanes, provides bi-directional bandwidth up to 8 GB / s.
For PCI Express 3.0, we will get a doubling of these indicators. PCI Express 3.0 uses a signaling rate of 8 GT / s, which provides 1 GB / s bandwidth per lane. Thus, the main slot for the video card will receive a bandwidth of up to 16 GB / s.
At first glance, the increase in signal speed from 5 GT / s to 8 GT / s does not seem like a doubling. However, PCI Express 2.0 uses an 8b / 10b encoding scheme, where 8 data bits are transmitted as 10-bit characters for error recovery. As a result, we get 20% redundancy, that is, a decrease in the usable bandwidth.
PCI Express 3.0 moves to a much more efficient 128b / 130b encoding scheme, eliminating 20% \u200b\u200bredundancy. Therefore 8 GT / s is no longer a "theoretical" speed; this is the actual rate, comparable in performance to the signal rate of 10 GT / s if the 8b / 10b coding principle were used.
 |
Click on the picture to enlarge.
We asked Yans about the devices that would require a speed boost. He replied that these will include "PLX switches, 40 Gbps Ethernet controllers, InfiniBand, the increasingly popular solid state devices, and of course graphics cards." He added “We haven't run out of innovations, they don't appear statically, they are a continuous flow,” they open the way for further improvements in future versions of the PCI Express interface.
Analysis: where will we use PCI Express 3.0?
Storage devices
AMD has already integrated SATA 6Gb / s support into its 8th line of chipsets, and motherboard manufacturers are adding USB 3.0 controllers. Intel is a little behind in this area, since it does not support USB 3.0 or SATA 6 Gb / s in chipsets (we already have preliminary samples of motherboards on the P67 in our laboratory, and they have support for SATA 6 Gb / s, but USB 3.0 in this generation we will not receive). However, as we have seen many times in the confrontation between AMD and Intel, AMD innovations often inspire Intel. Given the speed of the next generation storage interface and peripherals, there is no need to port any of the technologies to PCI Express 3.0 yet. For both USB 3.0 (5 Gb / s) and SATA 6 Gb / s (there are no drives yet that would come to the limits of this interface), one PCI Express second generation line will be enough.
Of course, when it comes to drives, the interaction between drives and controllers is only part of the question. Imagine an array of multiple SATA 6Gb / s SSDs on a chipset, where a RAID 0 array could potentially load one Gen 2 PCI Express lane that most motherboard manufacturers use to connect a controller. So, after some simple calculations, you can decide whether USB 3.0 and SATA 6 Gb / s interfaces can really require PCI Express 3.0 support.
 |
Click on the picture to enlarge.
As we already mentioned, the USB 3.0 interface gives a maximum speed of 5 Gbps. But also as a PCI Express 2.1 standard, USB 3.0 uses 8b / 10b encoding, which means that the actual peak speed is 4 Gbps. Divide the bits by eight to convert to bytes, and you get a peak throughput of 500MB / s - exactly the same as the single lane of the current PCI Express 2.1 standard. SATA 6 Gb / s runs at 6 Gb / s, but it also uses an 8b / 10b encoding scheme, which translates the theoretical 6 Gb / s into actual 4.8 Gb / s. Again, convert this value to bytes and you get 600 MB / s or 20% more than the PCI Express 2.0 line can provide.
However, the problem lies in the fact that even the fastest SSDs today cannot fully load a SATA 3Gb / s connection. The peripherals don't come close to the load of the USB 3.0 interface, the same can be said about the latest generation of SATA 6Gb / s. At least today, PCI Express 3.0 is not necessary for its active promotion on the platform market. Hopefully, as Intel moves to third-generation NAND flash, clock speeds will increase and we will have devices capable of surpassing the 3Gb / s level of second-generation SATA ports.
Video cards
We conducted our own research on the impact of PCI Express bandwidth on the performance of video cards - post-market PCI Express 2.0 , early 2010 and also recently ... We found it very difficult to load up the x16 bandwidth currently available on PCI Express 2.1 motherboards. You will need a multi-GPU configuration or an extreme high-end graphics card on a single GPU to spot the difference between x8 and x16 connections.
We asked AMD and Nvidia to comment on the need for PCI Express 3.0 - will this fast bus be required to unleash the full performance potential of next-gen graphics cards? An AMD spokesperson told us that he cannot comment yet.
 |
Click on the picture to enlarge.
An Nvidia spokesperson was more accommodating: “Nvidia has played a key role in the industry in developing PCI Express 3.0, which is supposed to double the bandwidth of the current generation (2.0) standard. Consumers and professionals will benefit from the new standard with increased graphics and compute performance in laptops, desktops, workstations and servers with GPUs. "
Perhaps the key phrase is "there will be applications that can use them." It looks like nothing is getting smaller in the graphics world. Displays are getting bigger, high definitions are replacing standard definition, textures in games are becoming more detailed and intriguing. Today we don't think that even the latest top-end graphics cards need a PCI Express 3.0 16-lane interface. But enthusiasts have seen history repeat itself year after year: advances in technology are paving the way for new ways to use "thicker pipes." Perhaps we will see an explosive growth in applications that will make GPU computing more massive. Or, perhaps, the drop in performance that is observed when the video card memory goes beyond the limits of memory, when swapping from the system memory starts, will no longer be so noticeable in mass and low-end products. In any case, we are about to see the innovations that PCI Express 3.0 will bring AMD and Nvidia.
Motherboard component connections
AMD and Intel are always very reluctant to share information about the interfaces they use to communicate between chipset components or logic "building blocks" in the north / south bridges. We know the speed at which these interfaces work, and also that they are designed so as not to create bottlenecks if possible. Sometimes we know who produced a certain part of the system logic, for example, AMD used a SATA controller in the SB600 based on the Silicon Logic design. But the technology used to build bridges between components often remains "blank spots." PCI Express 3.0, of course, seems like a very attractive solution, like the A-Link interface that AMD uses.
The recent appearance of USB 3.0 and SATA 6 Gb / s controllers on a large number of motherboards also allows us to assess the situation. Since the Intel X58 chipset does not provide native support for either of the two technologies, companies such as Gigabyte have to integrate controllers onto their motherboards using available lanes to connect them.
The Gigabyte EX58-UD5 motherboard does not support either USB 3.0 or SATA 6Gb / s. However, it does have an x4 PCI Express slot.
 |
Click on the picture to enlarge.
Gigabyte has replaced the EX58-UD5 motherboard with the new X58A-UD5, which has support for two USB 3.0 ports and two SATA 6Gb / s ports. Where did Gigabyte find the bandwidth to support these two technologies? The company took PCI Express 2.0 under one line for each controller, cutting down the possibilities for installing expansion cards, but at the same time enriching the functionality of the motherboard.
Apart from the addition of USB 3.0 and SATA 6Gb / s, the only noticeable difference between the two motherboards concerns the removal of the x4 slot.
 |
Click on the picture to enlarge.
Will the PCI Express 3.0 interface, as the standards before it, allow adding future technologies and controllers to motherboards that will not be present in the current generations of chipsets in an integrated form? As it seems to us, it will.
CUDA and parallel computing
We are entering the era of desktop supercomputers. Our systems run parallel-intensive GPUs, power supplies and motherboards capable of supporting up to four video cards simultaneously. Nvidia's CUDA technology transforms the graphics card into a tool for programmers to calculate not only in games, but also in scientific fields and engineering applications. The programming interface has already proven itself excellently when developing a variety of solutions for the corporate sector including medical imaging, mathematics, oil and gas exploration.
 |
Click on the picture to enlarge.
We asked for the opinion of OpenGL programmer Terry Welsh from Really Slick Screensavers about PCI Express 3.0 and GPU computing. Terry told us that “PCI Express has gotten a good head start and I love that developers double the bandwidth whenever they want - like with version 3.0. However, in the projects I have to work on, I don't expect to see any difference. my work is related to flight simulators, but they tend to be limited by memory and hard disk I / O performance; the graphics bus is not a bottleneck at all. But I can easily foresee that PCI Express 3.0 will provide significant progress for GPU computing; for people doing scientific work with big data. "
 |
Click on the picture to enlarge.
The ability to double the data rate while dealing with math-intensive workloads certainly motivates the development of CUDA and Fusion. And this is one of the most promising areas for the upcoming PCI Express 3.0 interface.
Any gamer with an Intel P55 chipset can talk about the advantages and disadvantages of the Intel P55 over the Intel X58 chipset. Advantage: Most P55 motherboards are more reasonably priced than Intel X58 models (overall, of course). Disadvantage: P55 has minimal PCI Express connectivity, the main task is assigned to Intel Clarkdale and Lynnfield processors, which have 16 PCIe lanes of the second generation in the CPU itself. Meanwhile, the X58 boasts 36 PCI Express 2.0 lanes.
For P55 buyers who wish to use two video cards, they will have to be connected via x8 lines each. If you want to add a third video card to the Intel P55 platform, you will have to use the chipset lines - but, unfortunately, they are limited by the speed of the first generation, and the chipset can allocate a maximum of four lines for the expansion slot.
When we asked Al Yans of PCI-SIG how many lanes to expect in chipsets supporting PCI Express 3.0 from AMD and Intel, he replied that this is "private information" that he "cannot disclose." Of course, we did not expect to receive an answer, but the question was still worth asking. However, it is unlikely that AMD and Intel, which are part of the PCI-SIG Board of Directors, would invest time and money in PCI Express 3.0 if they planned to use the new PCI Express standard simply as a means of reducing the number of lines. It seems to us that in the future AMD and Intel chipsets will continue to segment the way we see it today, high-end platforms will have enough options to connect a pair of video cards with a full x16 interface, and chipsets for the mass market will have cut lines.
Imagine a chipset like Intel P55, but with 16 PCI Express 3.0 lanes available. Since these 16 lanes are twice as fast as PCI Express 2.0, we get the equivalent of 32 lines of the old standard. In such a situation, it will be up to Intel to make the chipset compatible with 3-way and 4-way GPU configurations. Unfortunately, as we already know, the next generation Intel P67 and X68 chipsets will be limited to PCIe 2.0 support (and Sandy Bridge processors will be similarly limited to 16 lane support).
In addition to CUDA / Fusion parallel computing, we also see an increase in the capabilities of systems for the mass market due to the increased communication speed of PCI Express 3.0 components - here, too, we think there is a lot of potential hidden. Without a doubt, PCI Express 3.0 will improve the capabilities of low-cost motherboards that were only available to high-end platforms in the previous generation. And high-end platforms with PCI Express 3.0 at their disposal will allow us to set new performance records with innovations in graphics, storage subsystem and networking technologies that can use the available bus bandwidth.
