What are intel processors. Modern Intel and AMD processors

Even earlier than the Pentium MMX, the 6th generation processor, the Pentium Pro, appeared. In it, for the first time for IBM-compatible PCs, elements of the RISC architecture were used, which made it possible to increase productivity quite flexibly. However, the optimization of the processor for 32-bit programs and the high cost of production did not allow it to become mass-produced.

NOTE
Pentium Pro processors are classified as modern for the reason that the successor to the Pentium 4 - core processor 2 Duo was based on the Pentium Pro architecture, albeit deeply modernized.

Pentium II, Pentium III and Celeron

After making a number of improvements to the Pentium Pro and adding support for MMX instructions, Intel finally found a replacement for the Pentium and called it the Pentium II. The first Pentium IIs ran on a 66 MHz bus and had a native clock speed of 233 to 333 MHz. Then came the 100 MHz bus and new 350, 400 and 450 MHz processors. However new processor remained expensive for systems entry level, resulting in the Celeron - complete analogue Pentium II, except that it had less cache memory (and the first model did not have any) and worked only on the 66 MHz bus.

NOTE
Starting with the 386th processor, Intel began to use special, ultra-fast memory, located as close as possible to the processor. It stores data that is directly involved in the current calculation. This memory is called cache memory and significantly increases the speed of the PC. Its volume, as a rule, is from 128 to 512 KB.

The latest modification of the Pentium Pro is the Pentium III. It differs from its predecessor (Pentium II) primarily in the presence of SSE commands, which are much more efficient than MMX. The latest Pentium III and Celeron models operate at over 1 GHz.

Analogues: AMD Athlon(K7), AMD Duron.

Pentium 4

Late 2000 years Intel finally released the 7th generation processor. And while the Pentium 4 is the first processor that can't execute more instructions per cycle than its predecessor, it has very good potential for higher clock speeds. Already the first samples worked at 1.5 GHz (1500 MHz), and the latest models worked at clock frequency over 3.5 GHz, and Intel planned to release 10 GHz models by the end of 2010.

In addition to high clock speeds, the Pentium 4 has support for new SSE2 instructions designed to speed up video processing, and the latest models, starting at 3.06 GHz, can emulate the operation of two processors.

The features of the first systems based on the Pentium 4 include high power consumption - for stable operation it is recommended to use a power supply with a power of at least 300 W. The Pentium 4 was currently competing with AMD's Athlon XP and Ath-lon 64 processors.

Core 2 Duo, Core 2 Quad

Since serious technological and fundamental physical limitations prevented the release of processor models at frequencies of 4 GHz or more, in 2006 Intel released Core 2 family processors that could execute more instructions in one cycle and initially included 2 computing cores. Those. in fact, 2 full-fledged processors were placed in one crystal at once. A little later, 4-core (Core 2 Quad) models appeared. Thus, the gigahertz race was completed and the core race began.

Competitors - AMD Athlon X2, Phenom

Core i3/i5/i7

The latest Intel processors - Core i7 - inherited support for monothreading from Pentium 4, and from Core 2 - a high specific power of the computing core. So a 2-core Core i3/i5 has 4 virtual cores, a 4-core Core i7 has 8, and a 6-core Core i7 has a whopping 12!

Core i3/i5 competitors - AMD Athlon II/Phenom II X2/X3/X4, Core i7 - Phenom II X6.

I. How are processors measured?

There is one very serious problem of modern computer technology in general and processors in particular - how to independently and unambiguously evaluate the speed of a computer? Until recently, many compared the speed of processors with each other by their clock frequency. The vast majority of computer buyers thought that, for example, a computer with a "two thousand something" processor would be faster than "one thousand eight hundred." And the processor "two and a half thousand" is even faster. This was only partly true, because even then, in addition to "thousands of something there", processors had other characteristics: the system bus frequency - that is, the speed at which the processor "communicates" with the rest of the computer; cache size - that is, the size of the internal, private memory of the processor. For example, sometimes a 2.8 GHz Pentium IV processor with a 400 MHz system bus ran slower in some programs than a 2.6 GHz Pentium IV processor with a 533 MHz system bus. In this case, the natural frequency indicator - 2.8 GHz (or "two thousand eight hundred megahertz") was absolutely biased and not

displays the actual speed of the processor.

And now the situation has become even worse. Intel and AMD corporations began to increase the speed of their processors not at the expense of frequency, but at the expense of other parameters - internal circuitry and multi-core. That is, the first - the internal circuitry of the processor was significantly improved, due to which the processor began to process more information in one cycle of its work. Secondly, one of the most revolutionary decisions was made: instead of further straining with an increase in the speed of one processor, the engineers took and inserted two processors at once, or even four, into one physical microcircuit. Such solutions are called dual-core and quad-core. It is logical to assume that two processors together can process more information than one,

respectively, and they will work together faster. That, in general, and confirmed the practice. At a physical native processor speed of 1.86 GHz, for example, the new dual-core processors are several times faster than their previous 3.2 GHz or even 3.4 GHz Pentium IV brethren. At the same time, new processors heat up much less and consume much less electricity.

But how exactly do you compare the speed of new processors with old ones? How to measure the speed of work, in what units? If earlier, I repeat, people looked at the processor frequency (although this was not entirely true even then), then how can we evaluate the speed now? In what? It's so incomprehensible

to people uninitiated in these subtleties, that many are already beginning to invent some incomprehensible comparisons for themselves. For example, there is a statement that the processor frequency must be multiplied by the number of cores. Say, if dual core processor with a frequency of 1.86, which means each core runs at 1.86, which means the entire processor runs at 3.72. I'll tell you what - this is complete nonsense. People can’t understand that the processor’s weight entirely works at 1.86 GHz, and the speed is achieved due to a more advanced internal circuit and optimization of programs for multi-core, due to which its real speed with programs can be compared with the hypothetical Pentium IV 4 .5, or probably even with 5.0.

In order not to bother customers with all sorts of frequencies, caches and other characteristics, Intel Corporation has long ago made a logical marketing move - it introduced the processor number. Let me explain: each technical product has a certain model number, which accurately and unambiguously identifies a certain device with certain technical characteristics. And, it is quite logical that the higher this number, the newer the model and, accordingly, higher and better specifications. Entering the processor number greatly facilitates the choice of the necessary purchase for the buyer. Now you do not need to delve into the frequencies, caches, buses, now you just need to know the number (model) of the processor. All other things being equal, the processor, for example, the Core 2 Duo E8400 will be more powerful than the Core 2 Duo E7400. And you don't need to know that Core 2 Duo E7400 has a frequency of 2.8 GHz, 1066 MHz system bus, 3 MB cache, while Core 2 Duo E8400 has 3 GHz, 1333 MHz bus, 6 MB cache. You do not need to know all these figures, and even more so to understand them !!! It is enough to compare two numbers: 7400 and 8400. Well, and, of course, to see the difference in price.

And now let's look at what kind of processors our respected global manufacturers are producing today, in what cases and for what purposes these processors can be used.

II. Intel processors.
II.1 Why such diversity.
You know, I'll tell you a secret, at one of the Intel seminars for sellers, a company representative told us that Intel sets up all sellers to try to persuade buyers for the most powerful, latest and, of course, the most expensive processor models. In principle, this is correct, and the point here is not only that Intel is trying to simply increase profits in this way. The fact is that buying one of the fastest or even the fastest processor today, you get the most out of your computer and can perform the widest range of tasks.

But by conducting similar trainings for salespeople, Intel a little cunning, because at the same time it creates a huge range of absolutely new processors from the simplest and cheapest to the fastest and most expensive. Such a wide range of processors, which exists today at Intel, has never been in the history of this company.

Why is this happening? The fact is that most buyers of new computers probably know very little about them, or even know nothing about computers at all. But almost everyone has heard that computers are developing very quickly, almost every day they are becoming more powerful and more powerful. This is absolutely correct, but here's the thing: in recent years, computers have advanced so much, have developed so much that even the most inexpensive of the new modern computers easily cope with a very wide range of tasks.

Even if you take a computer based on the "weakest" of modern Intel processors - Celeron 430, then on such a computer you can easily perform any office work: a set of tests, essays, term papers, theses, you can prepare doctoral dissertations, you can surf the Internet, learn English and other languages, you can watch movies and listen to music, you can keep accounts of several enterprises. Why am I saying all this: when buying computers on very powerful and expensive processors Today, you may be overpaying for features that you most likely won't use.

That is why there is such a variety of processors. So that everyone can choose the most suitable computer both in terms of characteristics and price.

II.2 The lineup Intel processors.
If earlier all processors from Intel were divided into two large groups - Celeron and Pentium, then modern processors from Intel today can be divided into 4 large groups:

1. Celeron.
2. Pentium Dual Core.
3. Core 2 Duo.
4. Quadcore.

Just two more points about Intel: first, you may have a question: "Where did the Core 1 Duo processor or just Core Duo go? After all, if there is a Core 2 Duo, then in theory there should be the same processor, but without 2." That's right, such a processor exists, but it is produced only in special modifications for laptops, and such a processor does not exist for desktop computers. Secondly, in the price lists you can see a group of processors in the name of which there is the word Xeon. Ignore these processors, they exist for special powerful server computers designed to manage computer networks. In ordinary desktop computers Xeon processors do not apply.

III. AMD processors.
With the release of the K6 and K6-2 processors, AMD has become a full-fledged player in the microprocessor market. At first, AMD processors were thought to be cheap and fast enough. Then - as about inexpensive and the fastest. And when the price of AMD processors almost caught up with the price of processors from Intel, AMD had to think about cheap market segments. Emulating Intel with its Celeron processors, AMD began to release their processors with simplified features and inexpensive prices. These processors were called Duron. After a while these inexpensive processors became known as Sempron. To date, due to competition with Intel, AMD has had to significantly reduce the prices of its processors, as a result of which AMD's Athlon processors have become so cheaper that the need for even cheaper Semprons has completely disappeared. Athlon processors today they have occupied the niche of inexpensive products, but they have been replaced by more advanced and powerful processors phenom.

To date, AMD processors are divided into three large groups:

1. Athlon.
2. Phenom X3 - tri-core.
3. Phenom X4 - quad-core.

IV. Comparison and conclusions.
As you can see, today the prices for processors from AMD are significantly lower. What about speed? A very difficult question that I asked in the first chapter. How to measure the speed of two processors in the forehead and in what? There is a huge variety of test programs that are used to test various test laboratories of computer magazines. However, the results of these tests should only be partially trusted.

For example, if we run a test program on a computer based on Celeron, then the program starts working under the conditions of this particular computer, with the clock frequency of this particular processor, with this motherboard etc. That is, the program makes all measurements in some relative units relative to this particular computer. If the same program is run on a computer based on Core 2 Duo, then the program will take measurements in relative units of a faster computer.

Of course, the programmer tries to make the program independent of processors and computers, but this is quite difficult. Because, again, there are no single relative units of the speed of the processor in particular and the computer as a whole.

There are cases when a program is deliberately optimized by a programmer for one type of processor, for example, only for Intel or for AMD. And on a processor from another manufacturer, the program either does not work at all, or it works very slowly. That is why I would not recommend trusting various test programs, as well as the results of testing on these programs.

Subjectively, you can run several programs that you work with most often on several computers and visually compare how fast these programs will work. Thus, you can subjectively assess the speed of various computers.

In any case, you need to understand that the higher the processor model and, accordingly, its price, the faster the processor itself and the computer assembled on its basis. It remains for you to compare your needs from a computer with your financial capabilities and make the final choice.

Happy shopping!

Personal computer processors meet a single standard set by Intel, the world leader in the production of PC processors. In older computers, we can find processors of the Pentium II, Pentium III types, in the newest - Pentium 4. AMD produces processors that are generally similar to Intel ones, but they are called a little differently: K6 (second Pentium), K7 or Athlon (third Pentium). So AMD has to anticipate the future of the industry, sometimes outperforming Intel with its half-billion dollar revenue. Predictably, the emergence of new ideas from a lagging company is a way for it to survive. But it is surprising that sometimes these ideas are adopted by Intel. We are talking about IBM-compatible personal computers. In our market, as, indeed, in the world, they are the vast majority. Based on this standard, games, programs, and so on are written.

At the heart of any PC is the use of microprocessors. It is one of the most important devices in a computer, which is used to characterize the level of PC performance. The microprocessor is the "brain" and "heart" of the computer. It executes the programs running on the computer and controls the operation of other computer devices. When choosing a computer for themselves, the first thing they choose is a microprocessor that will meet the requirements of certain people. It depends on the processor how fast programs will run, and even how fast the process of archiving data in WinRAR will take place, not to mention the creation 3D animation in 3D MAX Studio. From all of the above, I believe that my topic is very relevant and significant today.

The purpose of my work is to compare several of the most popular processors to date and identify the leader among them.

Microprocessor - central device(or a complex of devices) a computer (or a computing system) that performs arithmetic and logical operations specified by the information conversion program, controls the computing process and coordinates the operation of the system devices (storage, sorting, input-output, data preparation, etc.). A computer system may have several parallel processors; such systems are called multiprocessor. The presence of several processors speeds up the execution of one large or several (including interconnected) programs. The main characteristics of the microprocessor are speed and capacity. Speed ​​is the number of operations performed per second. The bit depth characterizes the amount of information that the microprocessor processes in one operation: an 8-bit processor processes 8 bits of information in one operation, a 32-bit processor processes 32 bits. The speed of the microprocessor largely determines the speed of the computer. It performs all the processing of data entering the computer and stored in its memory, under the control of a program also stored in memory. Personal computers are equipped with central processors of various capacities.

Processor features:

data processing according to a given program by performing arithmetic and logical operations;

software control of the operation of computer devices.

Processor models include the following cooperating devices:

Control device (CU). It coordinates the work of all other devices, performs device management functions, and manages computer calculations.

Arithmetic logic unit (ALU). This is the name of the device for integer operations. Arithmetic operations such as addition, multiplication and division, as well as logical operations (OR, AND, ASL, ROL, etc.) are processed by the ALU. These operations make up the vast majority of the code in most programs. All operations in the ALU are performed in registers - specially designated cells of the ALU. A processor can have multiple ALUs. Each is capable of performing arithmetic or logical operations independently of the others, allowing multiple operations to be performed simultaneously. The arithmetic logic unit performs arithmetic and logical operations. Boolean operations are divided into two simple operations: "Yes" and "No" ("1" and "0"). Usually these two devices are allocated purely conditionally, they are not structurally separated.

AGU (Address Generation Unit) - address generation device. This device is no less important than the ALU, because. it is responsible for correct addressing when loading or saving data. Absolute addressing in programs is used only in rare exceptions. Once the data sets are taken, program code indirect addressing is used to make the AGU work.

Math coprocessor (FPU). A processor may contain several math coprocessors. Each of them is capable of performing at least one floating point operation regardless of what the other ALUs are doing. Data pipeline processing allows a single math coprocessor to perform multiple operations at the same time. The coprocessor supports high-precision calculations, both integer and floating point, and, in addition, contains a set of useful constants that speed up calculations. The coprocessor works in parallel with the central processor, thus providing high performance. The system executes coprocessor commands in the order they appear in the thread. Math coprocessor personal computer The IBM PC allows him to perform high-speed arithmetic and logarithmic operations, as well as trigonometric functions with high precision.

Decoder of instructions (commands). Parses instructions to extract operands and addresses where results are placed. This is followed by a message to another independent device about what needs to be done to execute the instruction. The decoder allows the execution of several instructions at the same time to load all the executing devices.

cache memory. Special high-speed processor memory. The cache is used as a buffer to speed up communication between the processor and RAM, and to store copies of instructions and data that have been recently used by the processor. Values ​​from the cache are retrieved directly without accessing the main memory. When studying the features of the work of programs, it was found that they access certain areas of memory with different frequencies, namely: memory cells that the program accessed recently are likely to be used again. Assume that the microprocessor is capable of storing copies of these instructions in its local memory. In this case, the processor will be able to use a copy of these instructions each time throughout the entire cycle. Access to memory is needed at the very beginning. A very small amount of memory is required to store these instructions. If the instructions to the processor arrive fast enough, then the microprocessor will not waste time waiting. This saves time for executing instructions. But for the fastest microprocessors, this is not enough. The solution to this problem is to improve the organization of memory. The memory inside the microprocessor can run at the speed of the processor itself.

First level cache (L1 cache). Cache memory located inside the processor. It is faster than all other types of memory, but smaller in size. Stores most recently used information that can be used in short program cycles.

Second level cache (L2 cache). Also located inside the processor. The information stored in it is used less frequently than the information stored in the first level cache, but it is larger in terms of memory.

Third level cache (L3 cache). Be inside the processor. The volume is larger than the memory of the first and second levels (512Kb-2Mb). Increases memory bandwidth.

main memory. Much larger than cache memory, and much slower.

Multi-level cache memory reduces the requirements of the most productive microprocessors to the speed of the main dynamic memory. So, if you reduce the access time to the main memory by 30%, then the performance of a well-designed cache memory will increase only by 10-15%. Cache memory, as you know, can have a rather strong impact on processor performance depending on the type of operations being performed, but increasing it does not necessarily bring an increase in overall processor performance. It all depends on how the application is optimized for this structure and uses the cache, and also on whether the various segments of the program are placed in the cache as a whole or in chunks.

The cache not only improves the speed of the microprocessor when reading from memory, but it can also store values ​​that the processor writes to main memory; it will be possible to write these values ​​later, when the main memory is not occupied. This cache is called a write-back cache. Its capabilities and principles of operation differ markedly from the characteristics of a write-through cache, which participates only in the read from memory operation.

A bus is a data transfer channel shared between different units in a system. The bus can be a set of conductive lines in a printed circuit board, wires soldered to the terminals of the connectors into which printed circuit boards or flat cable. Information is transmitted over the bus in the form of groups of bits. The bus may have a separate line for each bit of a word (parallel bus), or all the bits of a word may use the same line sequentially in time (serial bus). Many receiving devices - recipients - can be connected to the bus. Usually the data on the bus is destined for only one of them. The combination of control and address signals determines for whom. The control logic fires special strobe signals to indicate to the receiver when it should receive data. Receivers and senders can be unidirectional (ie only send or receive) or bidirectional (both). However, the fastest processor bus won't help much if the memory can't deliver the data at the appropriate rate.

Tire types:

Data bus. Serves to transfer data between the processor and memory or the processor and I / O devices. This data can be both microprocessor commands and information that it sends to or receives from I / O ports.

Address bus. Used by the CPU to select the required memory cell or I/O device by setting the bus to a specific address corresponding to one of the memory cells or one of the I/O elements included in the system.

Control bus. It carries control signals for memory and I/O devices. These signals indicate the direction of data transfer (to or from the processor).

BTB (Branch Target Buffer) - branch target buffer. This table contains all the addresses to which a transition will or can be made. Athlon processors also use a Branch History Table (BHT), which contains addresses that have already branched.

The registers are inner memory processor. They are a number of specialized additional memory cells, as well as internal storage media of the microprocessor. A register is a temporary storage device for data, a number, or an instruction, and is used to facilitate arithmetic, logic, and transfer operations. On the contents of some registers, special electronic circuits can perform some manipulations. For example, "cut" individual parts of the command for later use or perform certain arithmetic operations on numbers. The main element of the register is an electronic circuit called a flip-flop, which is capable of storing one binary digit (digit). The register is a collection of triggers related to each other in a certain way. common system management. There are several types of registers that differ in the type of operations performed.

Intel adheres to the EPIC (Explicitly Parallel Instruction Computing) standard. This technology was created specifically for large servers and some workstations. The possibilities of EPIC are enormous: firstly, it high speed perform floating point operations. Secondly, support for parallelization. And, thirdly, due to the improvement in reading data from memory, the speed of information exchange increases dramatically.

AMD has taken a different path to 64-bit. Manufacturers added 32 to the already existing digits and got new architecture x86-64. New technology differs from the old one only by the prefix 64. A number of improvements were made in the new processor, primarily the processor core. This made it possible to obtain a new level of performance for both 32 and 64-bit systems.

Bottom line: AMD moves to a new level without the use of new technologies. This results in full compatibility for both 32 and 64 bit applications. Intel tends to show itself only in 64 bits.

Big changes have been made to the new processors, which have resulted in performance and compatibility with older platforms.

AMD added compatibility modes and 64-bit address registers. They allow you to expand the addressable space random access memory and get rid of the existing limitation of 4 GB, which creates tangible difficulties in building information processing systems. To speed up the work with memory, NUMA technology is used, which allows you to work directly with memory, bypassing the system bus and chipset. This innovation was called HyperTransport and appeared in the first Golem chipset.

At Intel, things are much more complicated. Due to the intensive development path, the company has radically changed its architecture.

1. Modes of compatibility with old platforms.

2. Reducing the number of errors, since two independent technologies have been created against them. The main one is EMCA, which allows you to monitor and log all errors that occur during the operation of the processor. And a minor ECC technology that allows you to pre-process the code and maintain parity.

3. Support for multiprocessing. Since Intel oriented its processor for large servers, it also took care of multiprocessor. The processor was equipped with a number of microcircuits that allow for fast exchange with memory. Now, to work with the "brains", the methods of interleaving, buffering and dividing memory modules are used. At the same time, the processor works with 64 gigabytes of RAM with throughput 4.2 Gb/s

Intel has created a number of registers for full compatibility with older applications. As a result, it turns out that all 64-bit instructions are executed as usual, while others are processed by IA-32 technology. Emulation is emulation, no performance is involved, so Itanium is entirely 64-bit oriented.

In AMD, things are much more complicated. To improve performance with older platforms, special modes have been devised.

The AMD 64 architecture provides two main modes of operation: Long and Legacy. The first reveals all the advantages of x86-64 technology. For full compatibility over older applications, there is a compatibility submode that can process 32/16-bit instructions. In Legacy mode, the processor works on the principle of a conventional x86 architecture. The advantage of such a system of modes is that the processor can be operated until the release of stable releases of 64-bit operating systems. In addition, there are several advantages of x86-64 over IA-64:

1. Performance in processing 32-bit instructions. Due to the fact that after switching to compatibility mode, no emulation occurs, the processor processes data at a high speed. This is not the case in Itanium, since all instructions there are executed in 64 bits.

2. Full compatibility with x86 architecture. In Itanium, this is not fully implemented.

3. Simultaneous operation of 16/32/64 applications. Thanks to the introduction of modes, it becomes possible to process a number of different instructions simultaneously. This affects performance and improves compatibility.

Intel initially set itself the task of parallelizing processes in a single silicon device. As a rule, this processor is used on powerful servers with large databases or in banking systems where you cannot make mistakes. AMD, on the other hand, was oriented as a cross between 32 and 64 bits. Of course, it is found in large servers, but it can also be used in ordinary workstations, because it is tuned for both x86-64 and x86 architectures.

Intel asks no less than $1200 for its invention. And earlier the processor cost three times more: about $4k. Considering how much the motherboard for the processor will cost, we can conclude that a lot of money will have to be spent on the server.

At AMD price on Athlon 64 is only $417. The rest of the 64-bit processors cost between $300 and $600, well below Intel's prices.

The Celeron processor is a budget version of the corresponding main-stream processor, on the basis of whose core it was created. Celeron processors have two or four times less L2 cache. They also have a reduced system bus frequency compared to the corresponding "parents". Compared to Athlon, Duron processors have 4 times less cache memory and a lower 200MHz system bus (266MHz for Applebred), although there are "full-fledged" Athlons with 200MHz FSB. Cache-cut Bartons have also already appeared, the core of which is called Thorton. There are tasks in which there is almost no difference between regular and cut-down processors, and in some cases the lag is quite serious. On average, when compared with an uncut processor of the same frequency, the lag is 10-30%. But stripped-down processors tend to overclock better due to the smaller amount of cache memory and are cheaper at the same time. It should be noted that Celeron processors work very poorly compared to full-fledged P4 processors - the lag in some situations reaches 50%. This does not apply to processors. Celeron D, in which L2 cache is 256 KB (128 KB in regular Celerons) and the gap is not so big.

Firstly, AXP (and Athlon 64) have a rating instead of a frequency, i.e., for example, a 2000+ processor actually runs at 1667Mhz, but in terms of efficiency it corresponds to Athlon (Thunderbird) 2000Mhz. The main disadvantage has recently been considered temperature. But the latest models (based on Thoroughbred, Barton, etc.) are comparable to Pentium 4 in terms of heat dissipation, and the latest models from Intel (P4 Extreme Edition) at the time of writing the abstract sometimes heat up much more. In terms of reliability, processors are now also not much inferior to P4, although they cannot skip cycles when overheated, they have acquired a built-in thermal sensor. Athlon XP on the Barton core got a similar BusDisconnect function - it "disconnects" the processor from the bus during idle cycles, but it is virtually powerless in case of overheating from increased load - here all the "responsibility" is shifted to thermal control motherboard. Although the "strength" of the crystal has increased, it actually remained the same due to the reduced core area. Therefore, the probability of damage to the crystal, although it has become less, but exists. But in Athlon 64, the processor chip was finally hidden under the heat spreader, so it will be extremely difficult to damage it. All problems attributed to AMD are often the result of uninstalled or incorrectly installed universal drivers for VIA chipsets (VIA 4 in 1 Service Pack) or chipset drivers from other manufacturers (AMD, SIS, ALi).

However, many applications are not optimized and cannot benefit from dual or multi-core environments. To use multiple processors, software should be split into several parallel threads. This approach allows you to distribute the load across all available computing cores, reducing the computation time more than could be done using a single clock frequency. However, most programs today do not know how to use the capabilities of dual-core or multi-core chips.

Popular Dual Core AMD processors and Intel cost about $1,000, which is about the price of a complete computer. At the same time, single-core processors running at the same clock speed will cost only $300-$350.

For our comparison, professional-level processors were taken, namely: AMD Opteron and Intel Xeon. AMD is asking around $1,100 for a dual-core Opteron 275 (2.2GHz), while a pair of single-core Opteron 248s will cost just $700.

If you look at Intel, the situation is similar here. A dual-core 2.8GHz Xeon costs about $1,100, while two comparable 2.8GHz single-core Xeons cost about $550. Two 3.2GHz Xeons cost about $700.