What is pci express controller. What is PCI Express

If you ask which interface should be used for a solid-state drive with support for the NVMe protocol, then anyone (generally knowing what NVMe is) will answer: of course PCIe 3.0 x4! True, he will most likely have difficulties with justification. In the best case, we get the answer that such drives support PCIe 3.0 x4, and throughput interface matters. It has something, but all the talk about it began only when some drives became cramped for some operations within the framework of "regular" SATA. But between its 600 MB / s and (just as theoretical) 4 GB / s PCIe 3.0 x4 interface - there is simply a chasm, and filled with tons of options! What if one PCIe 3.0 lane is enough, since it is already one and a half times more than SATA600? Controller manufacturers who threaten to switch to PCIe 3.0 x2 in budget products, as well as the fact that many users do not have such and such, add fuel to the fire. More precisely, theoretically there is, but you can release them only by reconfiguring the system or even changing something in it, which you do not want to do. But buy a top-end solid state drive- I want to, but there are fears that there will be no benefit at all (even moral satisfaction from the results of test utilities).

But is it true or not? In other words, is it really necessary to focus exclusively on the supported mode of operation - or is it still possible in practice sacrifice principles? This is what we decided to check today. Let the check be quick and not claiming to be exhaustive, but the information received should be enough (as it seems to us) at least to think about ... But for now, let's briefly familiarize ourselves with the theory.

PCI Express: Existing Standards and Their Bandwidth

Let's start with what PCIe is and how fast it works. It is often called a "bus", which is somewhat incorrect ideologically: as such, there is no bus to which all devices are connected. In fact, there is a set of point-to-point connections (similar to many other serial interfaces) with a controller in the middle and attached devices (each of which itself could be a next-level hub).

The first version of PCI Express appeared almost 15 years ago. Orientation for use inside a computer (often within the same board) made it possible to make the standard high-speed: 2.5 gigatransactions per second. Because the interface is serial and duplex, a single PCIe lane (x1; actually an atomic unit) can transfer data at speeds up to 5 Gbps. However, in each direction - only half of this, that is, 2.5 Gbit / s, and this is the full interface speed, not "useful": to increase reliability, each byte is encoded with 10 bits, so the theoretical bandwidth of one PCIe line 1.x is approximately 250 MB / s each way. In practice, you still need to transfer service information, and as a result, it is more correct to talk about ≈200 MB / s of user data transfer. Which, however, at that time not only covered the needs of most devices, but also provided a solid supply: just remember that the predecessor of PCIe in the mass market segment system interfaces, namely the PCI bus, provided a bandwidth of 133 MB / s. And even if we consider not only the mass implementation, but also all PCI variants, the maximum was 533 MB / s, and for the entire bus, that is, such a PS was divided into all devices connected to it. Here, 250 MB / s (since for PCI, the full, and not useful, bandwidth is usually given) per line is in exclusive use. And for devices that need more, initially it was possible to aggregate several lines into a single interface, in powers of two - from 2 to 32, that is, the x32 option provided by the standard could transfer up to 8 GB / s in each direction. In personal computers, x32 was not used due to the complexity of creating and breeding the corresponding controllers and devices, so the option with 16 lines became the maximum. It was used (and is still used) mainly by video cards, since most devices do not need so much. In general, a considerable number of them and one line is quite enough, but some use both x4 and x8 with success: just on the storage topic - RAID controllers or SSDs.

Time did not stand still, and about 10 years ago the second version of PCIe appeared. Improvements concerned not only speeds, but in this respect a step forward was also taken - the interface began to provide 5 gigatransactions per second while maintaining the same encoding scheme, that is, the throughput has doubled. And once again it doubled in 2010: PCIe 3.0 provides 8 (not 10) gigatransactions per second, but the redundancy has decreased - now 130 bits are used for encoding 128 bits, not 160, as before. In principle, the PCIe 4.0 version with the next doubling of speeds is already ready to appear on paper, but in the near future we will hardly see it in hardware in large quantities. In fact, PCIe 3.0 is still used in conjunction with PCIe 2.0 in many platforms, because the performance of the latter is simply ... not needed for many applications. And where needed - the good old method of line aggregation works. Only each of them has become four times faster over the past years, that is, PCIe 3.0 x4 is PCIe 1.0 x16, the fastest slot in computers in the mid-2000s. This option is supported by top-end SSD controllers, and it is recommended to use it. It is clear that if there is such an opportunity, it is not enough. And if it is not there? Will there be any problems, and if so, what are they? It is with this question that we have to figure it out.

Testing methodology

Run tests with different versions PCIe standard is not difficult: almost all controllers allow using not only the supported by them, but also all earlier ones. It's more difficult with the number of lanes: we wanted to directly test the variants with one or two PCIe lanes. The board we usually use is the Asus H97-Pro Gamer on Intel chipset The H97 does not support the full set, but besides the "processor" x16 slot (which is usually used), it has another one that works in PCIe 2.0 x2 or x4 modes. We took advantage of this three, adding to it the PCIe 2.0 mode of the "processor" slot, in order to assess whether there is a difference. Still, in this case, there are no outside "intermediaries" between the processor and the SSD, but when working with a "chipset" slot - there is: the actual chipset, which is actually connected to the same PCIe 2.0 x4 processor. It was possible to add a few more modes of operation, but we were still going to carry out the bulk of the research on a different system.

The fact is that we decided to take advantage of the opportunity and at the same time check one "urban legend", namely the belief about the usefulness of using top-end processors for testing drives. So we took the eight-core Core i7-5960X - a relative of the Core i3-4170 usually used in tests (these are Haswell and Haswell-E), but which has four times more cores. In addition, the Asus Sabertooth X99 board found in the bins is useful for us today by the presence of a PCIe x4 slot, which in fact can work as x1 or x2. In this system, we tested three x4 options (PCIe 1.0 / 2.0 / 3.0) from the processor and chipset PCIe 1.0 x1, PCIe 1.0 x2, PCIe 2.0 x1 and PCIe 2.0 x2 (in all cases, chipset configurations are marked on the diagrams with a (c)). Does it make sense now to turn to the first version of PCIe, considering that there is hardly a single board supporting only this version of the standard that can boot from an NVMe device? From a practical point of view, no, but to check a priori the expected ratio of PCIe 1.1 x4 = PCIe 2.0 x2 and the like, it will be useful to us. If the check shows that the scalability of the bus is in line with theory, then it doesn't matter that we have not yet managed to get practically meaningful ways of connecting PCIe 3.0 x1 / x2: the first will be identical to PCIe 1.1 x4 or PCIe 2.0 x2, and the second - PCIe 2.0 x4 ... And we have them.

In terms of software, we limited ourselves to Anvil's Storage Utilities 1.1.0: it measures various low-level characteristics of drives well, but we don't need anything else. On the contrary: any influence of other components of the system is extremely undesirable, so that low-level synthetics for our purposes have no alternative.

We used a 240 GB Patriot Hellfire as a "working medium". As it was found during testing, it is not a record holder in performance, but its speed characteristics are quite consistent with the results best SSD of the same class and the same capacity. And there are already slower devices on the market, and there will be more and more of them. In principle, it will be possible to repeat the tests with something faster, however, as it seems to us, there is no need for this - the results are predictable. But let's not get ahead of ourselves, but let's see what we did.

Test results

While testing Hellfire, we noticed that the maximum speed on sequential operations can only be squeezed out of it with a multithreaded load, so this should also be taken into account for the future: the theoretical bandwidth is for that and the theoretical that the "real" data, received in different programs according to different scenarios, they will depend more not on it, but on these same programs and scenarios - in that case, of course, when force majeure circumstances do not interfere :) We are now observing such circumstances: it has already been said above that PCIe 1 .x x1 is ≈200 MB / s and this is what we see. Two PCIe 1.x lanes or one PCIe 2.0 is twice as fast, and this is exactly what we see. Four PCIe 1.x lanes, two PCIe 2.0 or one PCIe 3.0 are twice as fast, which was confirmed for the first two options, so the third is unlikely to differ. That is, in principle, scalability, as expected, is ideal: the operations are linear, Flash handles them well, so the interface matters. Flash stops do well to PCIe 2.0 x4 for recording (which means PCIe 3.0 x2 will do as well). Reading "may" is more, but last step gives already one and a half, and not two-fold (as it potentially should be) increase. We also note that there is no noticeable difference between the chipset and the processor controller, and between platforms, too. However, LGA2011-3 is slightly ahead, but only a little bit.

Everything is smooth and beautiful. But templates do not tear: the maximum in these tests is only slightly more than 500 MB / s, and this is quite capable of even SATA600 or (in the application to today's testing) PCIe 1.0 x4 / PCIe 2.0 x2 / PCIe 3.0 x1... That's right: don't be scared by the release of budget controllers for PCIe x2 or the presence of only such a number of lanes (and the version of the 2.0 standard) in M.2 slots on some motherboards, when more is not needed. Sometimes even that much is not needed: the maximum results were achieved with a queue of 16 teams, which is not typical for mass software. The queue with 1-4 commands is more common, and for this you can do with one line of the very first PCIe and even the very first SATA. However, there are overheads and so on, so the quick interface is useful. However, too fast is perhaps not harmful.

And in this test, platforms behave differently, and with a single command queue - in fundamentally different ways. The "problem" is not at all that many cores are bad. They are still not used here, except for one thing, and not so much that the boost mode unfolded with might and main. So we have a difference of about 20% in core frequency and one and a half times in cache memory - in Haswell-E it works at a lower frequency, and not synchronously with the cores. In general, the top-end platform can only be useful for knocking out the maximum of "yops" through the maximum multi-threaded mode with a large command queue depth. The only pity is that from the point of view practical work this is really spherical synthetics in a vacuum :)

On the record, the state of affairs has not fundamentally changed - in every sense. But, what is funny, on both systems the fastest was the PCIe 2.0 x4 mode in the "processor" slot. Both! And with multiple checks / rechecks. At this point you will inevitably think about whether you need these are your new standards or is it better not to rush anywhere at all ...

When working with blocks of different sizes, the theoretical idyll breaks down on the fact that increasing the interface speed still makes sense. The resulting figures are such that a pair of PCIe 2.0 lanes would be enough, but in reality, in this case, the performance is lower than that of PCIe 3.0 x4, albeit not by several times. And in general, the top-end platform "scores" here to a much greater extent. But just such operations are found mainly in application software, that is, this diagram is the closest to reality. As a result, it is not surprising that thick interfaces and fashionable protocols do not give any "wow effect". More precisely, those who pass over from the mechanics will be given, but exactly the same as any solid-state drive with any interface will provide.

Total

To facilitate the perception of the picture of the hospital as a whole, we used the score given by the program (total - for reading and writing), having carried out its standardization according to the "chipset" PCIe 2.0 x4 mode: at the moment it is he who is the most widely available, since it is found even on LGA1155 or AMD platforms without the need to "offend" the video card. In addition, it is equivalent to PCIe 3.0 x2, which budget controllers are preparing to master. Yes, and on the new AMD platform AM4, again, it is precisely this mode that can be obtained without affecting the discrete video card.

So what do we see? The use of PCIe 3.0 x4, if possible, is certainly preferable, but not necessary: ​​it brings literally 10% additional performance to NVMe mid-range drives (in its initially top-end segment). And even then - due to operations, in general, not so common in practice. Why, in this case, is this particular option implemented? Firstly, there was such an opportunity, and the stock does not hold a pocket. Secondly, there are drives that are faster than our test Patriot Hellfire. Thirdly, there are areas of activity where "atypical" for desktop system loads are just quite typical. And it is there that the performance of the data storage system is most critical, or at least the ability to make part of it very fast. But to the usual personal computers none of this applies.

As we can see, the use of PCIe 2.0 x2 (or, accordingly, PCIe 3.0 x1) does not lead to a dramatic decrease in performance - only by 15-20%. And this despite the fact that in this case we limited the potential capabilities of the controller by four times! For many operations, this bandwidth is sufficient. One PCIe 2.0 line is no longer enough, so it makes sense for controllers to support PCIe 3.0 - and in the face of a severe lack of lines in modern system this will work pretty well. Also, the x4 width is useful - even if there is no support modern versions PCIe in the system, it will still allow you to work at normal speed (albeit slower than it could potentially), if there is a more or less wide slot.

Basically, a large number of scenarios in which bottleneck it turns out that the flash memory itself (yes, this is possible and is inherent not only in mechanics), leads to the fact that four lanes of the third PCIe version on this drive overtake the first one by about 3.5 times - the theoretical throughput of these two cases differs by 16 once. From which, of course, it does not follow that you need to hurry to master very slow interfaces - their time has gone forever. It's just that many of the capabilities of fast interfaces can only be realized in the future. Or in the conditions with which regular user ordinary computer never in my life will directly face (except for those who like to measure it is known what). Actually, that's all.

). One of the criteria for this venture was the free home and household use of USB 3.0 ports. Well, since motherboards am2 platforms were not provided with such ports, then there is only one way out - to use the controller.
There were no questions about the delivery - the track was international and everything was tracked correctly.
But the packaging (or rather its absence) - discouraged. The seller is either naive to the point of recklessness or is the owner of a whole bunch of steel eggs. Because the antistatic bag with the controller just packs in the mail. No hint of packing / sealing material. As a result, the controller was delivered to me with a torn off capacitor (+ one more on parole). And in the comments through one, buyers write about a scratched disk with drivers or a bent mounting frame.

As for the driver disk. Specifically, in my case, I cannot check its serviceability / performance. Since I have no equipment to read it. But there are several nuances about this:
1) Under Windows 10, the controller does not require driver installation. (I confirm!)
2) In the reviews they write that the seller sends the driver upon request. (I cannot comment, I did not apply. There was no need)
3) The driver (for XP, VISTA, 7 and 8) can always be downloaded from the chip model - VL805.


Speaking frankly, except for the jamb with capacitors, there is no place to find fault with the installation. Soldered neatly and cleanly.
Geometrically, the board is small and can easily fit even into a skinny system unit, but there is no corresponding mounting frame for this. (Ideal, in my opinion, would be the choice when placing an order. But this is true, dreams ...)


The VL805 chip "knows how to make" four USB 3.0 ports. Two of which (A F) are directed outward, and two (19 pin) are inward system unit... Next to the latter is a 4-pin molex connector for external power supply.


For speed tests, I used the following external drives:
1) Memory card connected with via USB 3.0 card reader ().
[Further in all screenshots on the left there are measurements on the old port 2.0. And on the right on the new - 3.0]

2) 2.5 inch HDD Western Digital WD Scorpio Blue 320 GB () paired with USB 3.0 "pocket"

3) SSD OCZ Vertex 2 () paired with the aforementioned external box from AgeStar.

What can I say about this. The values ​​of the numbers turned out to be expected, because USB 2.0 is an old and slow port. And the difference in speed between it and the USB 3.0 port will be all the more noticeable, the faster the drive you intend to use.
If we talk specifically about this considered board, then due to the carelessness of the seller, it may very well be that when buying it you will have to work with a soldering iron and / or download a driver from the Internet. On the other hand, this is one of the cheapest options to equip your older PC with a faster interface.

Now I say goodbye. Be kind!