How to find out the frequency of the pci bus e. PCI interface in a computer: types and purpose. A photo. Information programs and utilities

“No one on this train knows anything!
“What else can you expect from these idle foreigners?”
Agatha Christie, Orient Express.

So, gentlemen, it's time to change the tire that has been the industry standard for 10 years. PCI, whose first version of the standard was developed back in 1991, has lived a long and happy life, in its various forms being the basis for small and large servers, industrial computers, laptops and graphic solutions (recall that AGP also traces its pedigree from PCI and is a specialized and extended version of the latter). But, before talking about the new product, let's look like historical attendants, remembering how the development of PCI took place. For, it has been repeatedly noted that, speaking about future prospects, it is always useful to find historical analogies: History of PCI

In 1991, Intel offers the base version (1.0) of the draft PCI (Peripheral Component Interconnect) bus standard. PCI is intended to replace ISA (and later its not very successful and expensive server extended modification EISA). In addition to the significantly increased throughput, the new bus is characterized by the ability to dynamically configure the resources (interrupts) allocated to connected devices.

In 1993, the PCI Special Interest Group (PCISIG, PCI Special Interest Group, the organization that took care of the development and adoption of various standards related to PCI) published an updated 2.0 revision of the standard that became the basis for the wide expansion of PCI (and its various modifications) in the information technology industry. Many well-known companies participate in PCISIG activities, including the ancestor of PCI Intel Corporation, which gave the industry many long-playing, historically successful standards. So, the basic PCI version (IEEE P1386.1):

• Bus clock 33 MHz, synchronous data transfer is used;
• Peak throughput 133 MB per second;
• Parallel data bus 32-bit wide;
• Address space 32-bit (4 GB);
• Signal level 3.3 or 5 volts.

Later, the following key tire modifications appear:

• PCI 2.2 allows 64-bit bus width and/or 66 MHz clock speed, i.e. peak throughput up to 533 MB/s;
• PCI-X, 64-bit version of PCI 2.2 with a frequency increased to 133 MHz (peak bandwidth 1066 MB/s);
• PCI-X 266 (PCI-X DDR), DDR version of PCI-X (effective frequency 266 MHz, real 133 MHz with transmission on both clock edges, peak bandwidth 2.1 GB/s);
• PCI-X 533 (PCI-X QDR), QDR version of PCI-X (effective frequency 533 MHz, peak bandwidth 4.3 GB/s);
• Mini PCI PCI with SO-DIMM style connector, mainly used for miniature network, modem and other cards in laptops;
• Compact PCI form factor standard (modules are inserted from the end into a cabinet with a common bus on the rear plane) and connector, designed primarily for industrial computers and other critical applications;
• Accelerated Graphics Port (AGP) high-speed PCI version optimized for graphics accelerators. There is no bus arbitration (i.e. only one device is allowed, with the exception of the latest, 3.0 version of the AGP standard, where there can be two devices and slots). Transfers towards the accelerator are optimized, there is a set of special additional features specific to graphics. For the first time this bus appeared together with the first system sets for the Pentium II processor. There are three basic versions of the AGP protocol, an additional specification for power (AGP Pro) and 4 data transfer rates from 1x (266 MB / s) to 8x (2Gb / s), including signal levels of 1.5, 1.0 and 0.8 volts.

We also mention CARDBUS 32-bit version of the bus for PCMCIA cards, with hot plugging and some additional features, nevertheless, having much in common with the basic version of PCI.

As we can see, the main development of the tire goes in the following directions:

1. Creation of specialized modifications (AGP);
2. Creation of specialized forms of factors (Mini PCI, Compact PCI, CARDBUS);
3. Increasing bit depth;
4. Increasing the clock frequency and the use of DDR / QDR data transfer schemes.

All this is quite logical, given the huge lifespan of such a universal standard. Moreover, points 1 and 2 do not aim to maintain compatibility with basic PCI cards, but points 3 and 4 are performed by increasing the original PCI slot, and allow the installation of conventional 32-bit PCI cards. In fairness, we note that during the evolution of the bus, there were also deliberate losses of compatibility with old cards, even for the basic version of the PCI connector, for example, in specification 2.3, the mention of support for 5 volt signal level and supply voltage disappeared. As a result, server boards equipped with this bus modification may suffer when old, five-volt cards are installed in them, although, in terms of connector geometry, these cards fit them.

However, like any other technology (for example, processor core architectures), bus technology has its own reasonable scaling limits, approaching which an increase in bandwidth comes at an increasing cost. An increased clock frequency requires more expensive wiring and imposes significant restrictions on the length of signal lines, an increase in bit depth or the use of DDR solutions also entails many problems, which in the end simply result in an increase in cost. And if in the server segment, solutions like PCI-X 266/533 will still be economically justified for some time, then we have not seen them in consumer PCs, and we will not see them. Why? Obviously, ideally, bus bandwidth should grow in sync with processor performance growth, while the implementation price should not only remain the same, but ideally also decrease. At the moment, this is only possible with the new bus technology. We will talk about them today: The era of serial buses

So, it's no secret that in this day and age, the ideal front-end is somehow consistent. Gone are the days of stranded centronics, and thick (you can't break a butt) SCSI hoses in fact, a heritage from before PC times. The transition happened slowly but surely: first the keyboard and mouse, then the modem, then, years and years later, scanners and printers, camcorders, digital cameras. USB, IEE1394, USB 2. At the moment, all consumer peripherals have moved to serial connections. Not far off and wireless solutions. The mechanism is obvious in our time it is more profitable to put maximum functionality into the chip (hot plugging, serial encoding, transmission and reception, data decoding, routing and error protection protocols, etc. necessary to squeeze out the necessary topological flexibility and significant bandwidth from a pair of wires things) rather than dealing with excessive volumes of contacts, hoses with hundreds of wires inside, expensive soldering, shielding, wiring and copper. Nowadays, serial buses are becoming more convenient not only from the point of view of the end user, but also from the point of view of the banal benefit bandwidth multiplied by distance divided by bucks. Of course, over time, this trend could not but spread to the inside of the computer we are already seeing the first fruit of this approach Serial ATA. Moreover, this trend can be extrapolated not only to the system buses (the main topic of this article) but also to the memory bus (it is fair to say that such an example has already been Rambus, but the industry rightly considered it premature) and even to the processor bus (potentially more good example HT). Who knows how many pins the Pentium X will have—maybe less than a hundred, assuming half of them are ground and power. Time to slow down and articulate the benefits of serial buses and interfaces:

1. Beneficial transfer of an increasing part of the practical implementation of the bus to silicon, which facilitates debugging, increases flexibility and reduces development time;
2. The prospect of organically using other signal carriers in the future, for example, optical ones;
3. Space saving (non-affordable miniaturization) and reduced installation complexity;
4. Easier to implement hot plugs and dynamic configuration in any sense;
5. Ability to allocate guaranteed and isochronous channels;
6. Transition from shared buses with arbitration and unpredictable interruptions inconvenient for reliable/critical systems to more predictable point-to-point connections;
7. Better cost and more flexible topology scalability;
8. Is this still not enough??? ;-).

In the future, we should expect a transition to wireless buses, technologies like UWB (Ultra Wide Band), however, this is not a matter of the next year or even five years.

And now, it's time to discuss all the benefits on a specific example the new standard PCI Express system bus, the mass distribution of which to the PC segment and medium / small servers is expected in the middle of next year. PCI Express just the facts

PCI Express Key Differences

Let's take a closer look at the key differences between PCI Express and PCI:

1. As has been repeatedly mentioned the new bus is serial, not parallel. Key benefits cost reduction, miniaturization, better scaling, more favorable electrical and frequency parameters (no need to synchronize all signal lines);
2. The specification is divided into a whole stack of protocols, each layer of which can be improved, simplified or replaced without affecting the others. For example a different signal carrier may be used or routing may be abolished in the case of a dedicated channel for only one device. Additional controls may be added. The development of such a bus will be much less painful increasing throughput will not require changing the control protocol and vice versa. Quickly and conveniently develop customized options for special purposes;
3. Hot-swappable cards were included in the original specification;
4. The initial specification includes the possibility of creating virtual channels, guaranteeing bandwidth and response time, collecting QoS statistics (Quality of Service Quality of Service);
5. The original specification included the ability to check the integrity of the transmitted data (CRC);
6. The original specification included power management capabilities.

So, wider ranges of applicability, more convenient scaling and adaptation, a rich set of originally incorporated features. Everything is so good that you just can't believe it. However, in relation to this tire, even inveterate pessimists speak out rather positively than negatively. And this is not surprising a candidate for the ten-year throne of the general standard for a large number of different applications (from mobile and embedded to enterprise-class servers or critical applications) simply has to look perfect from all sides, at least on paper :-). How it will be in practice we will soon see for ourselves. PCI Express what it will look like

The easiest way to switch to PCI-Express for standard desktop systems looks like this:

However, in the future it is logical to expect the appearance of a PCI Express splitter. Then the unification of the northern southern bridges will become quite justified. Let us give examples of possible system topologies. Classic PC with two bridges:

As already mentioned, a Mini PCI Express slot is provided and standardized:

And a new slot for external replaceable cards, similar to CARDBUS, which includes not only PCI Express but also USB 2.0:

Interestingly, there are two card form factors, but they differ not in thickness as before, but in width:

The solution is very convenient firstly, making a two-story installation inside the card is much more expensive and inconvenient than making a card with a larger area board inside, and secondly, a full-width card will end up with twice as much bandwidth, i.e. the second connector will not be idle. From an electrical or protocol point of view, the NewCard bus does not carry anything new, all the functions necessary for hot swapping or power saving are already included in the basic PCI Express specification.PCI Express transition

To facilitate the transition, a mechanism is provided for compatibility with software written for PCI (device drivers, OS). Besides, the PCI Express slots, unlike the PCI ones, are located on the other side of the section designated for the expansion card, i.e. can coexist in one place with PCI connectors. The user will only have to choose which card he wants to insert. First of all, PCI Express is expected to appear in Intel's entry-level server (dual-processor) platforms in the first half of 2004, followed by enthusiast-class desktop platforms and workstations (in the same year). How quickly PCI Express will be supported by other chipset manufacturers is not clear, however, both NVIDIA and SIS answer the question in the affirmative, although they do not name specific dates. Graphics solutions (accelerators) from NVIDIA and ATI equipped with built-in support for PCI Express x16 have long been planned and are being prepared for release in the first half of 2004. Many other manufacturers are active participants in the development and testing of PCI Express and also intend to present their products before the end of 2004.

Let's see! There is a suspicion that the baby came out successful.
Good luck, PCI Express: Departure 2004, Arrival 2014.

Current page: 6 (the book has 11 pages in total)

Font:

100% +

Chipset and bus overclocking options

By increasing the frequencies of the chipset and buses, you can increase their performance, however, in practice, it often becomes necessary to set these frequencies to fixed values ​​in order to avoid their excessive increase when overclocking the processor.

HT Frequency (LDT Frequency, HT Link Speed)

This parameter changes the frequency of the HT (HyperTransport) bus used by AMD processors with the chipset. Multipliers can be used as values ​​for this parameter, and the selected multiplier must be multiplied by the base frequency (200 MHz) to calculate the actual frequency. And in some versions of the BIOS, instead of multipliers, you need to select the HT bus frequency from several available values.

For processors of the Athlon 64 family, the maximum NT frequency was 800-1000 MHz (multiplier 4 or 5), and for Athlon P / Phenom II processors - 1800-2000 MHz (multiplier 9 or 10). When overclocking, the multiplier for the HT bus will sometimes have to be lowered so that after raising the base frequency, the HT frequency does not go beyond the permissible limits.

AGP/PCI Clock

This parameter sets the frequencies of the AGP and PCI buses.

Possible values:

□ Auto – frequencies are selected automatically;

□ 66.66/33.33, 72.73/36.36, 80.00/40.00 – AGP and PCI bus frequencies respectively. The default setting is 66.66/33.33, others can be used when overclocking.

PCIE Clock (PCI Express Frequency (MHz))

This parameter allows you to manually change the frequency of the PCI Express bus.

Possible values:

□ Auto – standard frequency is set (usually 100 MHz);

□ 90 to 150 MHz - the frequency can be set manually, and the adjustment range depends on the motherboard model.

CPU Clock Skew (MCH/ICH Clock Skew)

The parameters allow you to adjust the clock offset of the processor (CPU), as well as the north (MCH) and south (ICH) bridges.

Possible values:

□ Normal – the optimal value will be automatically set (recommended for normal operation and moderate overclocking);

□ 50 to 750 - amount of clock offset in picoseconds. Selecting this setting can improve system stability during overclocking.

FSB Strap to North Bridge

This parameter is used in some boards to set the operating mode of the chipset northbridge depending on the FSB frequency.

Possible values:

□ Auto – chipset parameters are configured automatically (this value is recommended for normal operation of the computer);

□ 200 MHz, 266 MHz, 333 MHz, 400 MHz – FSB frequency, for which the chipset operation mode is set. Higher values ​​increase the maximum possible FSB frequency during overclocking, but reduce the performance of the chipset. The optimal value of the parameter during overclocking usually has to be selected experimentally.

Chipset voltage adjustment

In addition to the processor and memory voltages, some motherboards also allow you to adjust the voltage of the chipset components and signal levels. The name of the corresponding parameters may be different depending on the board manufacturer. Here are some examples:

□ Chipset Core PCIE Voltage;

□ MCH & PCIE 1.5V Voltage;

□ PCH Core (PCH 1.05/1.8);

□ NF4 Chipset Voltage;

□ PCIE Voltage;

□ FSB OverVoltage Control;

□ NV Voltage (NBVcore);

□ SB I/O Power;

□ SB Core Power.

Practice shows that changing the specified voltages in most cases does not have a noticeable effect, so leave these voltages at Auto (Normal).

Spread spectrum

When the components of a modern computer operate at high frequencies, unwanted electromagnetic radiation occurs, which can be a source of interference for various electronic devices. In order to somewhat reduce the magnitude of the radiation pulses, spectral modulation of the clock pulses is used, which makes the radiation more uniform.

Possible values:

□ Enabled – the clock pulse modulation mode is enabled, which slightly reduces the level of electromagnetic interference from the system unit;

□ 0.25%, 0.5% – modulation level in percent (set in some BIOS versions);

□ Disabled - Spread Spectrum mode is disabled.

ADVICE

For stable system operation, always disable Spread Spectrum when overclocking.

Some motherboard models have several independent parameters that control the Spread Spectrum mode for individual system components, such as CPU Spread Spectrum, SATA Spread Spectrum, PCIE Spread Spectrum, etc.

Preparing for overclocking

Before overclocking, be sure to take a few important steps.

□ Check the stability of the system in normal mode. There is no point in overclocking a computer that is normally prone to crashes or freezes, as overclocking will only exacerbate this situation.

□ Find all the necessary BIOS settings that you will need for overclocking and understand their purpose. These parameters were described above, but for different models of boards they may vary, and to take into account the features of a particular board, you need to study the instructions for it.

□ Understand the BIOS reset method for your board model (see Chapter 5). This is necessary to reset the BIOS settings in case of unsuccessful overclocking.

□ Check the operating temperatures of the main components and their cooling. To monitor temperatures, you can use the diagnostic utilities from the CD-ROM to the motherboard or third-party programs: EVEREST, SpeedFan (www.almico.com), etc. To improve cooling, you may need to replace the CPU cooler with a more powerful one, and also take measures to improve the cooling of the chipset, video adapter and RAM.

Overclocking Intel Core 2 processors

The Intel Core 2 family of processors is one of the most successful in the history of the computer industry due to its high performance, low heat dissipation and excellent overclocking potential. Since 2006, Intel has released dozens of processors in this family under various brand names: Core 2 Duo, Core 2 Quad, Pentium Dual-Core and even Celeron.

To overclock Core 2 processors, you need to increase the FSB frequency, the nominal value of which can be 200, 266, 333 or 400 MHz. You can find out the exact value of the FSB frequency in the specification for your processor, but do not forget that the FSB frequency is indicated taking into account four times the multiplication during data transfer. For example, for the processor Core 2 Duo E6550 2.33 GHz (1333 MHz FSB), the actual value of the FSB frequency is 1333: 4 = 333 MHz.

Increasing the FSB frequency will automatically increase the operating frequencies of the RAM, chipset, PCI/PCIE buses, and other components. Therefore, before overclocking, you should forcibly reduce them in order to find out the maximum operating frequency of the processor. When it is known, you can choose the optimal operating frequencies for other components.

The sequence of acceleration can be as follows.

1. Set the optimal BIOS settings for your system. Select Disabled (Off) for Spread Spectrum, which is not very compatible with overclocking. You may have several such parameters: for the processor (CPU), PCI Express bus, SATA interface, etc.

2. Disable the Intel SpeedStep and C1E Support power saving technologies while overclocking. After all experiments are completed, you can enable these features again to reduce processor power consumption.

3. Set the PCI/PCIE bus frequencies manually. For the PCI bus, set the frequency to 33 MHz, and for PCI Express it is better to set the value within 100-110 MHz. On some board models, the Auto setting or the 100 MHz nameplate setting may result in worse results than the non-standard 101 MHz setting.

4. Reduce the frequency of the RAM. Depending on the board model, this can be done in one of two ways:

■ set the minimum value for the frequency of RAM using the Memory Frequency parameter or similar (to access this parameter, you may need to turn off automatic memory tuning);

■ set the minimum value of the multiplier that determines the ratio of the FSB frequency and memory using the FSB/Memory Ratio, System Memory Multiplier or similar parameter.

Since the ways to change the memory frequency vary between boards, it is recommended to restart the computer and use the EVEREST or CPU-Z diagnostic utilities to verify that the memory frequency has indeed decreased.

5. After the preparatory steps, you can proceed directly to the overclocking procedure. To begin with, you can raise the FSB frequency by 20-25% (for example, from 200 to 250 MHz or from 266 to 320 MHz), then try to load the operating system and check its operation. The parameter to set may be called CPU FSB Clock, CPU Overclock in MHz, or something else.

NOTE

To gain access to manual FSB adjustment, you may need to disable the automatic setting of the processor frequency (CPU Host Clock Control parameter) or dynamic overclocking of the motherboard. For example, on ASUS motherboards, set AI Overclocking (AI Tuning) to Manual.

6. Using the CPU-Z utility, check the actual operating frequencies of the processor and memory to make sure that your actions are correct (Fig. 6.3). Be sure to monitor operating temperatures and voltages. Run 1-2 test programs and make sure there are no crashes or freezes.

7. If the test of the overclocked computer was successful, you can restart it, increase the FSB frequency by 5 or 10 MHz, and then check the performance again. Continue until the system gives the first failure.

8. If a failure occurs, you can reduce the FSB frequency to return the system to a stable state. But if you want to know the maximum frequency of the processor, you need to increase the core voltage using the CPU VCore Voltage or CPU Voltage parameter. It is necessary to change the supply voltage smoothly and by no more than 0.1-0.2 V (up to 1.4-1.5 V). When testing a computer with an increased processor voltage, you should definitely pay attention to its temperature, which should not exceed 60 ° C. The final goal of this overclocking step is to find the maximum FSB frequency at which the processor can run for a long time without crashing and overheating.

9. Choose the optimal parameters of RAM. In step 4, we reduced its frequency, but as the FSB frequency increased, the memory frequency also increased. The actual value of the memory frequency can be calculated manually or determined using the utilities EVEREST, CPU-Z, etc. To speed up the memory, you can increase its frequency or reduce the timings, and to check the stability, use special memory tests: the MemTest utility or the built-in memory tests in diagnostic programs EVEREST and the like.

Rice. 6.3. Controlling the real frequency of the processor in the CPU-Z program

10. After the processor is overclocked and the optimal parameters of the memory bus are selected, you should comprehensively test the speed of the overclocked computer and the stability of its operation.

Overclocking of Intel Core i3/5/7 processors

Until 2010, Intel Core 2 processors were the most popular, but by that time, competing models from AMD had practically caught up with them in terms of performance and were also sold at lower prices. However, back in late 2008, Intel developed Core i7 processors with a completely new architecture, but they were produced in small batches and were very expensive. And only in 2010 is expected the arrival of chips with a new architecture to the masses. The company plans to release several models for all market segments: Core i7 - for productive systems, Core i5 - for the middle segment of the market and Core i3 - for entry-level systems.

The procedure for overclocking Intel Core i3/5/7 processors is not very different from overclocking Core 2 chips, but to get good results, you should take into account the main features of the new architecture: transferring the DDR3 memory controller directly to the processor and replacing the FSB bus with a new QPI serial bus. Similar principles have long been used in AMD processors, however, Intel has done everything at a very high level, and at the time of publication of the book, the performance of Core i7 processors is unattainable for competitors.

To set the operating frequencies of the processor, RAM, memory modules, DDR3 controller, cache memory and the QPI bus, the principle of multiplying the base frequency of 133 MHz (BCLK) by certain coefficients is used. Therefore, the main method of overclocking processors is to increase the base frequency, however, this will automatically increase the frequencies of all other components. As with Core 2 overclocking, you need to lower the RAM multiplier beforehand so that after increasing the base frequency, the memory frequency does not become too high. You may need to adjust the multipliers for the QPI bus and DDR3 controller under extreme overclocking, and in most cases these components will work fine at higher frequencies.

Based on the above, the approximate procedure for overclocking a system based on Core i3/5/7 can be as follows.

1. Set the optimal BIOS settings for your system. Disable Spread Spectrum, Intel SpeedStep and C1E Support, and Intel Turbo Boost Technology.

2. Set the minimum multiplier for RAM using the System Memory Multiplier or similar. In most boards, the minimum possible multiplier is 6, which corresponds to a frequency of 800 MHz in normal mode. ASUS motherboards use the DRAM Frequency parameter for this purpose, which should be set to DDR3-800 MHz.

3. After the preparatory steps, you can start increasing the base frequency using the BCLK Frequency parameter or similar. You can start with a frequency of 160-170 MHz, and then increase it stepwise by 5-10 MHz. As statistics show, for most processors it is possible to raise the base frequency to 180-220 MHz.

4. When the first failure occurs, you can slightly reduce the base frequency to return the system to a working state, and thoroughly test it for stability. If you want to get the most out of the processor, you can try to increase the supply voltage by 0.1-0.3 V (up to 1.4-1.5 V), but you should take care of more efficient cooling. In some cases, you can increase the overclocking potential of the system by raising the voltage of the QPI bus and the L3 (Uncore) cache memory, RAM, or the processor phase-locked loop system (CPU PLL).

5. After determining the frequency at which the processor can operate for a long time without failures and overheating, you can choose the optimal parameters for RAM and other components.

Overclocking AMD Athlon/Phenom Processors

In the mid-2000s, AMD produced quite good processors of the Athlon 64 family for that time, but the Intel Core 2 processors released in 2006 surpassed them in all respects. Released in 2008, Phenom processors never managed to catch up with Core 2 in terms of performance, and only in 2009 Phenom II processors were able to compete on equal terms with them. However, by this time, Intel already had a Core i7 ready, and AMD chips were used in entry-level and mid-level systems.

The overclocking potential of AMD processors is slightly lower than that of Intel Core, and depends on the processor model. The memory controller is located directly in the processor, and communication with the chipset is carried out via a special HyperTransport (HT) bus. The operating frequency of the processor, memory and HT bus is determined by multiplying the base frequency (200 MHz) by certain factors.

For overclocking AMD processors, the method of increasing the base frequency of the processor is mainly used, this will automatically increase the frequency of the HyperTransport bus and the memory bus frequency, so they will need to be reduced before overclocking. Also in the assortment of the company there are models with an unlocked multiplier (Black Edition series), and overclocking of such chips can be performed by increasing the multiplier; in this case, there is no need to adjust the parameters of the RAM and the NT bus.

You can overclock Athlon, Phenom or Sempron processors in the following order.

1. Set the BIOS settings that are optimal for your system. Disable Cool "n" Quiet and Spread Spectrum technologies.

2. Reduce the frequency of the RAM. To do this, you may first have to unset the memory parameters using SPD (Memory Timing by SPD or similar), and then specify the lowest possible frequency in the Memory Frequency for parameter or similar (Fig. 6.4).

3. Reduce the frequency of the HyperTransport bus using the HT Frequency parameter or similar (Fig. 6.5) by 1-2 steps. For example, for Athlon 64 processors, the nominal HT frequency is 1000 MHz (multiplier of 5) and you can lower it to 600-800 MHz (multiplier of 3 or 4). If your system has a parameter for setting the frequency of the memory controller built into the processor, such as CPU / NB Frequency, it is also recommended to reduce its value.

4. Set fixed frequencies for PCI (33 MHz), PCI Express (100-110 MHz) and AGP (66 MHz) buses.

5. After all the above actions, you can start overclocking itself. To begin with, you can raise the base frequency by 10-20% (for example, from 200 to 240 MHz), then try to load the operating system and check its operation. The parameter to set may be called CPU FSB Clock, CPU Overclock in MHz, or similar.

Rice. 6.4. Setting the RAM frequency

Rice. 6.5. Reducing the operating frequency of the HyperTransport bus

6. Using the CPU-Z utility, check the actual operating frequencies of the processor and memory. If the test of the overclocked computer passed without failures, you can continue to increase the base frequency by 5-10 MHz.

7. If a failure occurs, you can reduce the base frequency to return the system to a stable state, or continue overclocking by increasing the core voltage (Fig. 6.6). You need to change the supply voltage smoothly and by no more than 0.2-0.3 V. When testing a computer with an increased processor supply voltage, pay attention to the processor temperature, which should not exceed 60 ° C.

Rice. 6.6. Increasing the processor core voltage

8. After overclocking the processor, set the optimal frequency for the NT bus, RAM and its controller, test the speed and stability of the overclocked computer. To reduce processor heat, enable Cool "n" Quiet technology and check the stability of work in this mode.

Unlocking cores in Phenom ll/Athlon II processors

The AMD Phenom II processor family, which was released in 2009, has various models with two, three, and four cores. Dual- and triple-core models were released by AMD by disabling one or two cores in a quad-core processor. This was explained by considerations of economy: if a defect was found in one of the cores of a quad-core processor, it was not thrown away, but the defective core was turned off and sold as a three-core one.

As it turned out later, a locked core can be enabled using the BIOS, and some of the unlocked processors can work normally with all four cores. This phenomenon can be explained by the fact that over time, there were fewer defects in the production of quad-core processors, and since there was a demand on the market for two- and three-core models, manufacturers could forcibly turn off completely working cores.

At the time of the publication of the book, it was known about the successful unlocking of most models of this family: Phenom II X3 series 7xx, Phenom II X2 series 5xx, Athlon II X3 series 7xx, Athlon II X3 series 4xx and some others. In the quad-core Phenom II X4 8xx and Athlon II X4 6xx models, there is a possibility of unlocking the L3 cache, and in the single-core Sempron 140 - the second core. The probability of unlocking depends not only on the model, but also on the batch in which the processor was released. There were parties in which it was possible to unlock more than half of the processors, and in some parties only rare instances could be unlocked.

To unlock, the motherboard BIOS must support Advanced Clock Calibration (ACC) technology. This technology is supported by AMD chipsets with the SB750 or SB710 southbridge, as well as some NVIDIA chipsets, such as GeForce 8200, GeForce 8300, nForce 720D, nForce 980.

The unlocking procedure itself is simple, you just need to set the Auto value for the Advanced Clock Calibration parameter or similar. In some boards from MSI, the Unlock CPU Core option should also be enabled. In case of failure, you can try to set up the ACC manually by experimentally choosing the value of the Value parameter. Sometimes, after turning on the ACC, the system may not boot at all, and you will have to reset the CMOS content using a jumper (see Chapter 5). If by no means you managed to unlock the processor, disable ACC, and the processor will work normally.

You can check the parameters of an unlocked processor using the EVEREST or CPU-Z diagnostic utilities, but to make sure that the result is positive, you should conduct a comprehensive computer test. Unlocking is done on the motherboard and does not change the physical state of the processor. You can refuse to unlock at any time by disabling ACC, and when you install the unlocked processor on another board, it will again be blocked.

In this article, we will explain the reasons for the success of the PCI bus and describe the high-performance technology that is coming to replace it - the PCI Express bus. We will also look at the history of development, the hardware and software levels of the PCI Express bus, the features of its implementation and list its advantages.

When in the early 1990s it appeared, then, in terms of its technical characteristics, it significantly exceeded all the tires that existed up to that moment, such as ISA, EISA, MCA and VL-bus. At that time, the PCI bus (Peripheral Component Interconnect - interaction of peripheral components), operating at a frequency of 33 MHz, was well suited for most peripheral devices. But today the situation has changed in many ways. First of all, the clock speeds of the processor and memory have increased significantly. For example, the clock frequency of processors has increased from 33 MHz to several GHz, while the operating frequency of PCI has increased to only 66 MHz. The emergence of technologies such as Gigabit Ethernet and IEEE 1394B threatened that the entire bandwidth of the PCI bus could go to serve a single device based on these technologies.

At the same time, the PCI architecture has a number of advantages over its predecessors, so it was not rational to completely revise it. First of all, it does not depend on the type of processor, it supports buffer isolation, bus mastering technology (bus capture) and PnP technology in full. Buffer isolation means that the PCI bus operates independently of the internal processor bus, which allows the processor bus to function independently of the speed and load of the system bus. With bus capture technology, peripheral devices can directly control the process of transferring data on the bus, instead of waiting for help from the central processor, which would affect system performance. Finally, Plug and Play support allows automatic configuration and configuration of devices using it and avoids fuss with jumpers and switches, which pretty much ruined the lives of owners of ISA devices.

Despite the undoubted success of PCI, at the present time it faces serious problems. Among them are limited bandwidth, lack of real-time data transmission functions and lack of support for next-generation network technologies.

Comparative characteristics of various PCI standards

It should be noted that the actual throughput may be less than the theoretical one due to the principle of the protocol and the features of the bus topology. In addition, the total bandwidth is distributed among all devices connected to it, therefore, the more devices sit on the bus, the less bandwidth goes to each of them.

Such standard improvements as PCI-X and AGP were designed to eliminate its main drawback - low clock speed. However, increasing the clock frequency in these implementations has resulted in a reduction in the effective length of the bus and the number of connectors.

The new generation of the bus, PCI Express (or PCI-E for short), was first introduced in 2004 and was designed to solve all the problems that its predecessor faced. Today, most new computers are equipped with a PCI Express bus. Although they also have standard PCI slots, the time is not far off when the bus will become history.

PCI Express Architecture

The bus architecture has a layered structure as shown in the figure.

The bus supports the PCI addressing model, which allows all currently existing drivers and applications to work with it. In addition, the PCI Express bus uses the standard PnP mechanism provided by the previous standard.

Consider the purpose of the various levels of organization PCI-E. At the software level of the bus, read / write requests are generated, which are transmitted at the transport level using a special packet protocol. The data layer is responsible for error-correcting coding and ensures data integrity. The basic hardware layer consists of a double simplex channel consisting of a transmit and receive pair, collectively referred to as a link. The total bus speed of 2.5 Gb/s means that the throughput for each PCI Express lane is 250 Mb/s each way. If we take into account the overhead costs of the protocol, then about 200 Mb / s is available for each device. This bandwidth is 2-4 times higher than what was available for PCI devices. And, unlike PCI, if the bandwidth is distributed among all devices, then it goes to each device in full.

To date, there are several versions of the PCI Express standard, which differ in their bandwidth.

PCI Express x16 bus bandwidth for different PCI-E versions, Gb/s:

• 32/64
• 64/128
• 128/256

PCI-E bus formats

At the moment, various options for PCI Express formats are available, depending on the purpose of the platform - a desktop computer, laptop or server. Servers that require more bandwidth have more PCI-E slots, and those slots have more trunks. In contrast, laptops may only have one line for medium-speed devices.

Video card with PCI Express x16 interface.

PCI Express expansion cards are very similar to PCI cards, but the PCI-E connectors are more grippy to ensure the card won't slip out of the slot due to vibration or during shipping. There are several form factors of PCI Express slots, the size of which depends on the number of lanes used. For example, a bus with 16 lanes is referred to as PCI Express x16. Although the total number of lanes can be as high as 32, in practice, most motherboards nowadays are equipped with a PCI Express x16 bus.

Smaller form factor cards can be plugged into larger form factor slots without compromising performance. For example, a PCI Express x1 card can be plugged into a PCI Express x16 slot. As in the case of the PCI bus, you can use a PCI Express extender to connect devices if necessary.

The appearance of connectors of various types on the motherboard. From top to bottom: PCI-X slot, PCI Express x8 slot, PCI slot, PCI Express x16 slot.

Express Card

The Express Card standard offers a very simple way to add hardware to a system. The target market for Express Card modules are laptops and small PCs. Unlike traditional desktop expansion cards, the Express card can be connected to the system at any time while the computer is running.

One of the popular varieties of Express Card is the PCI Express Mini Card, designed as a replacement for Mini PCI form factor cards. A card created in this format supports both PCI Express and USB 2.0. PCI Express Mini Card dimensions are 30×56 mm. PCI Express Mini Card can connect to PCI Express x1.

Benefits of PCI-E

PCI Express technology has gained advantages over PCI in the following five areas:

1. Better performance. With just one lane, the throughput of PCI Express is twice that of PCI. In this case, the throughput increases in proportion to the number of lines in the bus, the maximum number of which can reach 32. An additional advantage is that information can be transmitted along the bus in both directions simultaneously.
2. Simplification of input-output. PCI Express takes advantage of buses such as AGP and PCI-X while offering a less complex architecture and relatively simple implementation.
3. Layered architecture. PCI Express offers an architecture that can adapt to new technologies without the need for significant software upgrades.
4. New generation I/O technologies. PCI Express gives you new opportunities to receive data with the help of simultaneous data transfer technology, which ensures that information is received in a timely manner.
5. Ease of use. PCI-E greatly simplifies system upgrades and expansions by the user. Additional Express card formats, such as the ExpressCard, greatly increase the ability to add high-speed peripherals to servers and laptops.

Conclusion

PCI Express is a bus technology for connecting peripherals, replacing technologies such as ISA, AGP, and PCI. Its use significantly increases the performance of the computer, as well as the user's ability to expand and update the system.

Practical overclocking of the processor

Processor Overclocking Methods

There are two methods of overclocking "a: increasing the frequency of the system bus (FSB) and increasing the multiplier (multiplier). At the moment, the second method cannot be applied to almost all serial AMD processors. Exceptions to the rule are: Athlon XP processors (Thorubbred, Barton, Thorton )/Duron (Applebred) released before week 39, 2003, Athlon MP, Sempron (socket754; downgrade only), Athlon 64 (downgrade only), Athlon 64 FX53/55. In Intel production processors, the multiplier is also fully locked. by increasing the multiplier is the most "painless" and simplest, because only the processor clock frequency increases, and the frequencies of the memory bus, AGP / PCI buses remain nominal, so determine the maximum processor clock frequency at which it can work correctly using this It is a pity that it is rather difficult, if at all, to find AthlonXP processors with an unlocked multiplier for sale now. Maybe. Overclocking a processor by increasing the FSB has its own characteristics. For example, with an increase in the FSB frequency, the frequency of the memory bus and the frequency of the AGP/PCI buses increase. Particular attention should be paid to the PCI/AGP bus frequencies, which are related to the FSB frequency in most chipsets (does not apply to nForce2, nForce3 250). This dependence can be bypassed only if the BIOS of your motherboard has the appropriate parameters - the so-called dividers, which are responsible for the ratio of PCI / AGP to FSB. You can calculate the divider you need using the FSB / 33 formula, i.e., if the FSB frequency = 133 MHz, then you should divide 133 by 33, and you will get the divider you need - in this case it is 4. The nominal frequency for the PCI bus is 33 MHz, and the maximum is 38-40 MHz, it is not recommended to set it higher, to put it mildly: this can lead to the failure of PCI devices. By default, the memory bus frequency rises synchronously with the FSB frequency, so if the memory does not have enough potential for overclocking, it can play a limiting role. If it is obvious that the frequency of the RAM has reached its limit, you can do the following:

• Increase memory timings (for example, change 2.5-3-3-5 to 2.5-4-4-7 - this can help you squeeze a few more MHz out of RAM).
• Increase the voltage on the memory modules.
• Overclock CPU and memory asynchronously.

Reading is the mother of learning

First you need to study the instructions for your motherboard: find the sections of the BIOS menu that are responsible for the FSB frequency, RAM, memory timings, multiplier, voltages, PCI / AGP frequency dividers. If the BIOS does not have any of the above parameters, then overclocking can be done using jumpers (jumpers) on the motherboard. You can find the purpose of each jumper in the same instructions, but usually information about the function of each is already printed on the board itself. It happens that the manufacturer himself deliberately hides "advanced" BIOS settings - to unlock them, you need to press a certain key combination (this is often found on Gigabyte motherboards). I repeat: all the necessary information can be found in the instructions or on the official website of the motherboard manufacturer.

Practice

We go into the BIOS (usually, to enter, you need to press the Del key at the time of recalculating the amount of RAM (that is, when the first data appeared on the screen after restarting / turning on the computer, press the Del key), but there are models of motherboards with a different key for entering the BIOS - for example, F2), we are looking for a menu in which you can change the frequency of the system bus, memory bus and control timings (usually these parameters are located in one place). I think that overclocking the processor by increasing the multiplier will not cause difficulties, so let's move on to raising the system bus frequency. Raise the FSB frequency (by about 5-10% of the nominal), then save the changes, reboot and wait. If everything is fine, the system starts up with the new FSB value and, as a result, with a higher processor (and memory, if you overclock them synchronously) clock speed. Booting Windows without any excesses means that half the battle is already done. Next, we launch the CPU-Z program (at the time of this writing, its latest version was 1.24) or Everest and make sure that the processor clock speed has increased. Now we need to check the processor for stability - I think everyone has a 3DMark 2001/2003 distribution kit on the hard drive - although they are designed to determine the speed of the video card, you can "drive" them to superficially check the stability of the system. For a more serious test, you need to use Prime95, CPU Burn-in 1.01, S&M (more on testers below). If the system has been tested and behaves stably, we reboot and start all over again: go into the BIOS again, increase the FSB frequency, save the changes and test the system again. If during testing you were "thrown out" of the program, the system hung or rebooted, you should "roll back" a step back - to the processor frequency when the system behaved stably - and conduct more extensive testing to make sure the work is completely stable. Don't forget to keep an eye on the processor temperature and PCI/AGP bus frequencies (you can check the PCI frequency and temperature in the OS using the Everest program or motherboard manufacturer's proprietary programs).

Voltage boost

It is not recommended to increase the voltage on the processor by more than 15-20%, but it is better that it varies within 5-15%. There is a sense in this: the stability of work increases and new horizons for overclocking open up. But be careful: as the voltage rises, the power consumption and heat dissipation of the processor increase, and as a result, the load on the power supply increases and the temperature rises. Most motherboards allow you to set the voltage on the RAM to 2.8-3.0 V, the safe limit is 2.9 V (to further increase the voltage, you need to make a volt mod on the motherboard). The main thing when increasing the voltage (not only on RAM) is to control heat generation, and, if it has increased, organize cooling of the overclocked component. One of the best ways to determine the temperature of any computer component is to touch it with your hand. If you cannot touch a component without pain from a burn, it needs urgent cooling! If the component is hot, but you can hold your hand, then cooling would not hurt it. And only if you feel that the component is barely warm or even cold, then everything is fine, and it does not need cooling.

Timings and frequency dividers

Timings are delays between individual operations performed by the controller when accessing memory. There are six of them: RAS-to-CAS Delay (RCD), CAS Latency (CL), RAS Precharge (RP), Precharge Delay or Active Precharge Delay (more commonly referred to as Tras), SDRAM Idle Timer or SDRAM Idle Cycle Limit, Burst Length . Describing the meaning of each is meaningless and useless to anyone. It is better to immediately find out what is better: small timings or high frequencies. There is an opinion that timings are more important for Intel processors, while frequency is more important for AMD. But do not forget that for AMD processors, the memory frequency achieved in synchronous mode is most often important. For different processors, "native" are different memory frequencies. For Intel processors, the following combinations of frequencies are considered "friendly": 100:133, 133:166, 200:200. For AMD on nForce chipsets, synchronous operation of FSB and RAM is better, and asynchrony has little effect on the AMD + VIA bundle. On systems with an AMD processor, the memory frequency is set in the following percentages with FSB: 50%, 60%, 66%, 75%, 80%, 83%, 100%, 120%, 125%, 133%, 150%, 166% , 200% - these are the same divisors, but presented a little differently. And on systems with an Intel processor, the dividers look more familiar: 1:1, 4:3, 5:4, etc.

Black screen

Yes, it also happens :) - for example, when overclocking: you just set such a clock frequency of the processor or RAM (perhaps you specified too low memory timings) that the computer cannot start - or rather, it starts, but the screen remains black, and the system does not give any "signs of life". What to do in this case?

• Many manufacturers build in their motherboards a system for automatically resetting parameters to nominal values. And after such an “incident” with an overestimated frequency or low timings, this system should do its “dirty” work, but this does not always happen, so you need to be ready to work with the handles.
• After turning on the computer, press and hold the Ins key, after which it should start successfully, and you must go into the BIOS and set the computer's operating parameters.
• If the second method does not help you, you need to turn off the computer, open the case, find the jumper on the motherboard that is responsible for resetting the BIOS settings - the so-called CMOS (usually located near the BIOS chip) - and set it to Clear CMOS mode for 2-3 seconds, and then return to the nominal position.
• There are models of motherboards without a BIOS reset jumper (the manufacturer relies on its automatic BIOS reset system) - then you need to remove the battery for a while, which depends on the manufacturer and model of the motherboard (I conducted such an experiment on my Epox EP-8RDA3G: I took out the battery, waited 5 minutes, and the BIOS settings were reset).

Information programs and utilities

CPU-Z is one of the best programs that provides basic information about the processor, motherboard and RAM installed in your computer. The program interface is simple and intuitive: there is nothing superfluous, and all the most important things are in plain sight. The program supports the latest innovations from the hardware world and is updated periodically. The latest version at the time of this writing is 1.24. Size - 260 Kb. You can download the program at cpuid.com.

Everest Home/Professional Edition (formerly AIDA32) is an information and diagnostic utility that has more advanced functions for viewing information about the installed hardware, operating system, DirectX, etc. The differences between the home and professional versions are as follows: the Pro version does not have a RAM test module (read / write), it also lacks a rather interesting Overclock subsection, which contains basic information about the processor, motherboard, RAM, processor temperature, motherboard board and hard drive, as well as overclocking your processor as a percentage :). The Home version does not include software accounting, advanced reports, database interaction, remote management, and enterprise-level functions. In general, this is all the differences. I myself use the Home version of the utility, because I don't need the extra features of the Pro version. I almost forgot to mention that Everest allows you to view the PCI bus frequency - to do this, expand the Motherboard section, click on the subsection with the same name and find the Chipset bus properties / Real frequency item. The latest version at the time of this writing is 1.51. The Home version is free and weighs 3 Mb, the Pro version is paid and takes 3.1 Mb. You can download the utility at lavalys.com.

Stability testing

The name of the CPU Burn-in program speaks for itself: the program is designed to "warm up" the processor and check its stable operation. In the main CPU Burn-in window, you need to specify the duration, and in the options, select one of two testing modes:

• testing with enabled error checking;
• testing with disabled error checking, but with maximum "warming up" of the processor (Disable error checking, maximum heat generation).

When the first option is enabled, the program will check the correctness of the processor's calculations, and the second will allow the processor to "warm up" to temperatures close to the maximum. CPU Burn-in weighs about 7 Kb.

The next decent CPU and RAM tester is Prime95. Its main advantage is that when an error is detected, the program does not spontaneously hang up, but displays data about the error and the time it was detected on the working field. By opening the Options -> Torture Test… menu, you can choose from three testing modes yourself or specify your own parameters. For more effective detection of processor and memory errors, it is best to set the third testing mode (Blend: test some of everything, lots of RAM tested). Prime95 is 1.01 Mb and can be downloaded from mersenne.org.

Relatively recently, the S&M program saw the light. At first, it was conceived to check the stability of the processor power converter, then it was implemented to check the RAM and support for Pentium 4 processors with HyperThreading technology. At the moment, the latest version of S&M 1.0.0(159) supports more than 32 (!) processors and checks the stability of the processor and RAM, in addition, S&M has a flexible system of settings. Summarizing all of the above, it can be argued that S&M is one of the best programs of its kind, if not the best. The program interface is translated into Russian, so it is quite difficult to get confused in the menu. S&M 1.0.0(159) weighs 188 Kb and can be downloaded from testmem.nm.ru .

The above tester programs are designed to check the stability of the processor and RAM and identify errors in their work, they are all free. Each of them loads the processor and memory almost completely, but I want to remind you that programs used in everyday work and not intended for testing can rarely load the processor and RAM so much, so we can say that testing takes place with a certain margin.

The author does not bear any responsibility for the breakdown of any hardware of your computer, as well as for failures and "glitches" in the operation of any software installed on your computer.

#PCI

Attention! This article is about the PCI bus and its PCI64 and PCI-X derivatives! Do not confuse it with the newer tire ("PCI Express"), which is completely incompatible with the tires described in this FAQ.

PCI 2.0- the first version of the basic standard, which was widely used, both cards and slots with a signal voltage of only 5V were used.

PCI 2.1- differed from 2.0 by the possibility of simultaneous operation of several bus-master devices (the so-called competitive mode), as well as the appearance of universal expansion cards capable of operating in both 5V and 3.3V slots. The ability to work with 3.3V cards and the presence of appropriate power lines in version 2.1 was optional. PCI66 and PCI64 extensions appeared.

PCI 2.2- a version of the basic bus standard that allows connection of expansion cards with a signal voltage of both 5V and 3.3V. The 32-bit versions of these standards were the most common slot type at the time the FAQ was written. 32-bit, 5V type slots are used.
Expansion cards made in accordance with these standards have a universal connector and are able to work in almost all later varieties of PCI bus slots, and also, in some cases, in 2.1 slots.

PCI 2.3- the next version of the common standard for the PCI bus, expansion slots that comply with this standard are not compatible with PCI 5V cards, despite the continued use of 32-bit slots with a 5V key. Expansion cards have a universal connector, but are not able to work in 5V slots of earlier versions (up to 2.1 inclusive).
We remind you that the supply voltage (not signal!) 5V is stored absolutely on all versions of the PCI bus connectors.

PCI 64- an extension of the basic PCI standard, introduced in version 2.1, doubling the number of data lines, and, consequently, the throughput. The PCI64 slot is an extended version of the regular PCI slot. Formally, the compatibility of 32-bit cards with 64-bit slots (subject to the presence of a common supported signal voltage) is complete, and the compatibility of a 64-bit card with 32-bit slots is limited (in any case, there will be a loss of performance), exact data in each specific case can be found in the specifications of the device.
The first versions of PCI64 (derived from PCI 2.1) used a 64-bit 5V PCI slot and ran at 33MHz.

PCI 66- an extension of the PCI standard that appeared in version 2.1 with support for a clock frequency of 66 MHz, as well as PCI64, allows you to double the bandwidth. Starting with version 2.2, it uses 3.3V slots (the 32-bit version is almost never found on a PC), cards have a universal or 3.3V form factor. (There were also solutions based on version 2.1, casuistically rare on the PC 5V 66MHz market, such slots and boards were only compatible with each other)

PCI 64/66- A combination of the above two technologies, it can quadruple the data transfer rate compared to the basic PCI standard, and uses 64-bit 3.3V slots, compatible only with universal and 3.3V 32-bit expansion cards. PCI64/66 cards have a universal (with limited compatibility with 32-bit slots) or 3.3V form factor (the latter option is fundamentally not compatible with 32-bit 33MHz slots of popular standards)
Currently, the term PCI64 means exactly PCI64/66, since 33MHz 5V 64-bit slots have not been used for a long time.

PCI-X 1.0- Expansion of PCI64 with the addition of two new operating frequencies, 100 and 133 MHz, as well as a separate transaction mechanism to improve performance when running multiple devices at the same time. Generally backwards compatible with all 3.3V and universal PCI cards.
PCI-X cards are usually made in 64-bit 3.3 format and have limited backward compatibility with PCI64/66 slots, and some PCI-X cards are in a universal format and can work (although this has almost no practical value) in regular PCI 2.2 /2.3.
In difficult cases, in order to be completely confident in the performance of the combination of motherboard and expansion card you have chosen, in the case you need to look at the compatibility lists of the manufacturers of both devices.

PCI-X 2.0- further expansion of the capabilities of PCI-X 1.0, added speeds of 266 and 533 MHz, as well as parity error correction during data transfer. (ECC). It allows splitting into 4 independent 16-bit buses, which is used exclusively in embedded and industrial systems, the signal voltage is reduced to 1.5V, but the connectors are backward compatible with all cards using a 3.3V signal voltage.

PCI-X 1066/PCI-X 2133- projected future versions of the PCI-X bus, with resulting operating frequencies of 1066 and 2133 MHz, respectively, originally intended for connecting 10 and 40 Gbit Ethernet adapters.

For all variants of the PCI-X bus, there are the following restrictions on the number of devices connected to each bus:
66MHz - 4
100MHz - 2
133MHz - 1 (2, if one or both devices are not on expansion boards, but are already integrated on one board along with the controller)
266.533MHz and above -1.

That is why in some situations, in order to ensure the stability of several installed devices, it is necessary to limit the maximum frequency of the used PCI-X bus (usually this is done by jumpers)

CompactPCI- a standard for connectors and expansion cards used in industrial and embedded computers. Mechanically not compatible with any of the "common" standards.

MiniPCI- a standard for boards and connectors for integration into laptops (usually used for wireless network adapters) and directly to the surface. It is also mechanically incompatible with anything other than itself.

Types of PCI expansion cards:

Summary table of constructs of cards and slots depending on the version of the standard:

Summary table of compatibility of cards and slots depending on the version and design:

	Cards
Slots	PCI 2.0/2.1 5B	PCI 2.1 generic	PCI 2.2/2.3 universal	PCI64/5B (33MHz)	PCI64/universal	PCI64/3.3B	PCI-X/3.3B	PCI-X universal
PCI 2.0	Compatible	Compatible	Incompatible		Limited compatibility with performance loss	Incompatible
PCI 2.1	Compatible	Compatible	Limited compatible	Limited compatibility with performance loss	Limited compatibility with performance loss	Incompatible
PCI 2.2	Compatible			Limited compatibility with performance loss	Limited compatibility with performance loss	Incompatible	Incompatible	Limited compatibility with performance loss
PCI 2.3	Incompatible	Limited compatible	Compatible	Incompatible	Limited compatibility with performance loss	Incompatible	Incompatible	Limited compatibility with performance loss
PCIB 64/5B(33MHz)	Compatible	Compatible	Limited compatible	Compatible	Limited compatibility with performance loss	Incompatible	Incompatible	Limited compatibility with performance loss
PCI64/3.3B	Incompatible	Limited compatible	Compatible	Incompatible	Compatible	Compatible	Limited compatibility with performance loss	Limited compatibility with performance loss
PCI-X	Incompatible	Limited compatible	Compatible	Incompatible	Compatible