Decryption of AMD processors. Marking and overclocking potential of processors

Choosing a processor is quite a serious task, which should be approached only after you have thoroughly familiarized yourself with all the nuances and characteristics. Much can be learned from the name of the processor, its markings, which contain information about the main characteristics of this model. What these characteristics mean is possible, and in this article we will talk about how to decipher the processor markings.

Intel processor markings

Intel processor series
- I7– top processors that support all Intel technologies, have 4 cores, and are equipped with an 8 MB L3 cache memory.
- I5– mid-price segment processors can have from 2 to 4 cores. Equipped with L3 cache memory with a capacity of 3 to 6 MB. There is no support for Trusted Execution, Hyper-Threading and Virtualization Technology.
- I3– a budget series of processors, has 2 cores and an L3 cache with a capacity of 3 MB.
Indicates the generation of the processor series Core i-x. SandyBridje is marked with the number 2, IvyBridge is marked with the number 3.
Indicates position in the series. The higher the number, the faster the processor runs. Depends on the clock frequency.
Processor version
- K– such a processor has an unlocked multiplier, which means it can be overclocked.
- M– processor used in mobile devices (smartphone, tablet).
- P– processor without automatic overclocking.
- S– such processors have reduced power consumption to 65 W.
- T– these processors have reduced power consumption to 45/35 W.

AMD processor markings

Processors without a GPU video core.

Indicates the processor series.
Talks about the number of cores in the processor.
Indicates the processor architecture: number 2 – Bulldozer , 3 – Piledriver.
Determines the position of the model in the family; in most cases, it depends on the processor clock speed.

Processors with a built-in GPU video core.

Talks about the number of processor cores and the presence of a GPU video core.
- A10– there are 4 CPU cores and a Radeon HD 7660D video core (here and below for the Trinity architecture).
- A8— 4 CPU cores and a Radeon HD 7560D video core.
- A6— 2 CPU cores and a Radeon HD 7540D video core are available.
- A4— 2 CPU cores and a Radeon HD 7480D video core are available.
Indicates the processor generation.
This marking depends on the frequency; the higher the frequency, the greater the value.

The labeling of AMD Athlon 64 processors is so complex and confusing that it confuses even experts.

Model numbers, or ratings, were invented by AMD back when it was struggling to compete with Intel. The idea behind the processor pseudo-clock is to explain to the user what kind of Intel processor they are buying. Even then, AMD began to promote the thesis that processor performance depends not only on the clock frequency, but also on other parameters, primarily on the microarchitecture and the size of the built-in cache memory. The processor rating (“pseudo-frequency”) takes into account the difference in other parameters and can be used to compare processors from different manufacturers. And AMD began assigning “plus” numbers to its processors, indicating the clock speed of Intel processors of similar performance.

The starting point for the CPU rating was indeed calculated based on benchmark results. However, AMD then began to assign ratings simply in ascending order. And when different variants of the Athlon 64 appeared, the situation with ratings completely got out of control: now, using the rating, it was necessary to note not only the difference in frequencies, but also different caches, different processor sockets, etc. Therefore, in the summary table you will find many processors with the same numbers, but different parameters. They can be distinguished only by the marking line (OPN), which is printed on the processor case directly under its name.

When AMD introduced a new manufacturing technology, it decided not to change either the name or the way the processors were labeled. And therefore, Athlon 64 with different core revisions, differing in support for different frequencies and types of memory, instruction set support, power consumption and overclocking potential, can only be distinguished by OPN. Fortunately, the Athlon 64 boxed packaging has a transparent window through which the processor markings can be easily read. And don’t accidentally buy a processor with an old core, which may not support DDR400 or cannot be overclocked at all.

Note that AMD has been starting to improve lately. Among the processors based on the latest modification of the Venice core, there are no longer models with the same numbers and different parameters. We can already talk about the unique correspondence of the number to the frequency and cache size. Let's say, 3200+ will always have a frequency of 2 GHz and a cache of 512 KB, and only two numbers are “reserved” for processors with a 1 MB cache - 3700+ and 4000+.

Especially for overclocking fans, we will inform you that processors based on the Venice core (E3, E6), regardless of the number, are usually overclocked to 2.8-2.9 GHz. Therefore, it makes sense to buy the most affordable model - 3000+, since it, with proper luck, will allow you to reach the theoretical limit of its core.

Decoding processor numbers and markings of Athlon 64

Name- tion	Number	Marking	Core	Hour- tota	L2 cache	Tire	Those- pro- cess	Socket
AEPAP: processors based on the ClawHammer core (130 nm)*
Athlon 64	2800+	ADA2800AEP4AP	C0	1.8 GHz	0.5 MB	x4	130 nm	Socket 754
	3000+	ADA3000AEP4AP		2 GHz
	3200+	ADA3200AEP5AP		2 GHz	1 MB
	3400+	ADA3400AEP5AP		2.2 GHz
AEPAX/AR: processors based on a stripped-down NewCastle core*
Athlon 64	2800+	ADA2800AEP4AR	C.G.	1.8 GHz	0.5 MB	x4	130 nm	Socket 754
		ADA2800AEP4AX
	3000+	ADA3000AEP4AR		2 GHz
		ADA3000AEP4AX
	3200+	ADA3200AEP4AX		2.2 GHz
		ADA3200AEP5AR		2 GHz	1 MB
	3400+	ADA3400AEP4AR		2.4 GHz	0.5 MB
		ADA3400AEP4AX
		ADA3400AEP5AR		2.2 GHz	1 MB
	3700+	ADA3700AEP5AR		2.4 GHz	1 MB
AI4BX: processors based on a stripped-down Venice core (90 nm)*
Athlon 64	3000+	ADA3000AIK4BX	E6	2 GHz	0.5 MB	x4	90 nm	Socket 754
	3200+	ADA3200AIO4BX	E6	2.2 GHz
	3400+	ADA3400AIK4BO	E3	2.4 GHz
DEPA: NewCastle core processors (130 nm, dual-channel memory controller)
Athlon 64	3000+	ADA3000DEP4AW	C.G.	1.8 GHz	0.5 MB	x5	130 nm	Socket 939
	3200+	ADA3200DEP4AW		2 GHz
	3500+	ADA3500DEP4AS		2.2 GHz
		ADA3500DEP4AW
	3800+	ADA3800DEP4AS		2.4 GHz
		ADA3800DEP4AW
	4000+	ADA4000DEP5AS			1 MB
DIK4BI: Winchester core processors (90 nm, 512 KB cache)
Athlon 64	3000+	ADA3000DIK4BI	D0	1.8 GHz	0.5 MB	x5	90 nm	Socket 939
	3200+	ADA3200DIK4BI		2 GHz
	3500+	ADA3500DIK4BI		2.2 GHz
DAA4BP: Venice core processors (E3)
Athlon 64	3000+	ADA3000DAA4BP	E3	1.8 GHz	0.5 MB	x5	90 nm	Socket 939
	3200+	ADA3200DAA4BP		2 GHz
	3500+	ADA3500DAA4BP		2.2 GHz
	3800+	ADA3800DAA4BP		2.4 GHz
DAABN: processors based on SanDiego core (cache up to 1 MB)*
Athlon 64	3500+	ADA3500DAA4BN	E4	2.2 GHz	0.5 MB	x5	90 nm	Socket 939
	3700+	ADA3700DAA5BN		2.2 GHz	1 MB
	4000+	ADA4000DAA5BN		2.4 GHz
DAA4BW: Venice core processors (512 KB cache)
Athlon 64	3000+	ADA3000DAA4BW	E6	1.8 GHz	0.5 MB	x5	90 nm	Socket 939
	3200+	ADA3200DAA4BW		2 GHz
	3500+	ADA3000DAA4BW		2.2 GHz
	3800+	ADA3000DAA4BW		2.4 GHz
DKACG/CF: low power processors*
Athlon 64	3200+	ADA3200DKA4CG	E4	2 GHz	0.5 MB	x5	90 nm	Socket 939
	3500+	ADA3500DKA4CG		2.2 GHz	0.5 MB
	3700+	ADA3700DKA5CF	E6	2.2 GHz	1 MB
	4000+	ADA4000DKA5CF		2.4 GHz	1 MB

Those users who have just started overclocking have probably paid attention to the statistics on overclocking of a particular computer component, given in the press or on forums. The interest in this topic is understandable: one’s own experiments do not always give the same result as described on specialized sites. In order not to rely on luck when choosing a central processor, you should study the factors on which the overclocking potential depends.

First, let's figure out why overclocking becomes possible at all. The fact is that none of the modern manufacturers have separate lines for the production of processors with a given clock frequency. CPUs are simply produced using a certain technology (say, 0.09-micron), and already at the moment they “come off the assembly line” the fun begins. Take AMD Athlon 64 for example: processors based on the Venice core are manufactured at the Fab30 factory in Dresden. During the “assembly” process, testing is carried out, not of the entire batch, but only of its part. The test is first carried out at the maximum frequency for a given core (for example, 2.4 GHz), and if the sample taken passes all the tests, then all processors from this series are marked and released for sale with the appropriate rating (in this case, Athlon 64 3800+) . If a failure occurs during testing, the processor clock speed is lowered and the next testing cycle begins, and so on until the test batch passes all the procedures, after which the devices are assigned the appropriate rating.

The fact that only a part of the processors are tested allows us to assume that the model labeled as 3000+ (1800 MHz) will be able to operate at a frequency of 2400 MHz (which corresponds to 3800+), or even higher. It also happens that at the moment the market is oversaturated with expensive processors, but there is a shortage of weak and cheap ones, then the manufacturer makes a move and sells the older model under the guise of a younger one, lowering its clock frequency (and previously the available cache memory was also limited).

But how can we determine when buying a processor what it is actually capable of? An absolutely accurate answer can only be obtained through testing, but there are entire portals on the Internet where statistics on overclocking processors are given, indicating their markings. Very useful information can be indicated here, including the place and week of production of a given processor instance. How to read such a marking? To do this you need to know how it is decrypted.

So, the main thing that interests us in the labeling of AMD processors is the name of the core and the release date of this instance. And this is why we need such information: over time, from the start of production of a processor on a given core, the technology is debugged, i.e., there is less “rejection”, which means there is less chance of getting a copy with impressive overclocking. An example of this is the modern Athlon 64 on the Winchester core, which at first pleased enthusiasts with its very good potential, but after the 50th week of 2004 it barely reaches 2400 MHz.

Thus, knowing the overclocking statistics for AMD processors will help you make the right choice. Athlon 64 existed and exists on various cores - all of them are given in the description of the markings, and therefore we will not repeat them. Of all the listed options, only Winchester, Venice, San Diego for Athlon 64 processors, Paris and Palermo for Sempron are currently available for sale.

Winchester

Athlon 64 3000+ (1800 MHz) processors based on the Winchester core are gradually disappearing from store shelves. Once one of the most popular processors, it has now passed the baton to its successor, the Venice. The most successful release dates for processors based on this core are considered to be 43, 49 and 50 weeks of 2004. Instances produced during these weeks almost always confidently reached a frequency of 2500 MHz, and some even 2700–2750 MHz. However, it is impossible to find a well-overclocked processor released after the 50th week of 2004. Whether because AMD began to more carefully select copies for marking, or for another reason, it is now almost impossible to overcome the 2500 MHz threshold. Although overclocking is still a lottery, anything can happen here, for example, the 3rd week of 2005 often delighted overclockers with frequencies of 2600–2650 MHz. The worst are considered to be the 11th week of 2005 – overclocking to the level of 2200–2250 MHz – and the 1st week of 2005 – 2250–2300 MHz.

Venice

The new processor core, which replaced the public's favorite Winchester, has already become the talk of the town: a perfectly overclockable processor, a new 0.09 micron process technology (as opposed to 0.13 for Winchester). Venice repeats the feat of its great-grandfather Barton (Athlon XP Socket A). The new technical process (correspondingly, a reduced core area) made it possible to raise the frequency potential bar even higher - to the level of 2700–2800 MHz; it is rare that a processor based on this core does not reach the 2700 MHz mark. The most successful are the 16th, 20th, 22nd, 28th and 29th weeks of 2005. Almost all processors produced at that time are overclocked to 2750 MHz, and on average even to 2800 MHz. It is worth paying special attention to 22 and 28 weeks, this is when copies were produced that operate at a frequency of 2900–2950 MHz!

The worst weeks of 2005 were 17, 18 and 19. Overclocking rarely reaches 2600 MHz, which, to put it mildly, is not enough for this core. The limit for week 17 is 2550–2600 MHz, for week 18 – 2500–2550, while not a single processor from week 19 was able to reach the 2500 MHz bar.

San Diego

Another new core that makes a somewhat contradictory impression: it is, in principle, the same Venice, but with the second level cache increased to 1 MB. The San Diego core is used for the Athlon FX-57 processors, which is the reason for the weak overclocking potential of the Athlon 64 with such a core: processors are very meticulously selected for the Athlon FX, so it is unlikely to find a successful copy among the Athlon 64 San Diego.

The best production time was the 15th week of 2005: overclocking to the level of 2850–2900 MHz, the worst was the 22nd week of 2005 – overclocking to only 2500–2600 MHz.

Sempron

Sempron processors are designed for the budget segment of the market, they have a reduced second-level cache size - only 256 KB, do not support AMD64 technology, only older models have SSE3. Although there are now processors with a rating of 3300+ for Socket 939 that have 64-bit support, most likely this is a rejection of the Athlon 64 on the Venice core with the second level cache half disabled.

Paris

Processors based on this core do not please us with anything special: overclocking to the level of 2200–2500 MHz. The most successful production time was week 4 of 2005, overclocked at about 2450–2500 MHz. The worst one is week 37, 2004, 2200–2250 MHz.

Palermo

The transition to the 0.09 micron process technology did not give any particular advantage to Sempron processors based on the Palermo core - the same stable overclocking to the level of 2250–2500 MHz. The most successful copies were released in the 1st week of 2005 - overclocked to 2450-2550 MHz, the most unsuccessful - in the 5th and 27th weeks of 2005 - overclocked to about 2100-2150 MHz.

Athlon 64 FX

Dual-core processors of the new generation are produced on Manchester and Toledo cores, the difference is in the volume of the second level cache - 2x512 KB and 2x1024 KB, respectively. The products are not yet very popular due to their high cost, their production technology has not been fully developed, there are not many programs that get at least some performance boost from dual-core technology, and there are no games at all. Overclocking shows modest figures of the order of 2200–2400 MHz, but there are no representative statistics yet due to the “dampness” and lack of demand of these processors.

So, a small “population census” has been completed among modern processors produced by AMD - the most popular among enthusiasts due to their low cost, good performance and excellent overclocking potential. But it’s too early to put an end to it, new data is coming in, new records are being set. The beginning of 2006 is just around the corner, when processors based on the 0.065-micron process technology will appear and new overclocking opportunities will open up.

In general, it’s worth recalling that statistics are a very capricious thing, and even more so overclocking statistics. It is not a fact that by purchasing a processor from the best week according to statistics, you can achieve the results presented in this article. Overclocking depends on many factors, but the main ones in this difficult task are good cooling and a certain amount of luck.

AMD processor markings

ADA3000AEP4AX (first line)
LBBE 0529CPCW (second line)

These letters mean (for the first line):

ADA – processor type:
- ADA – Athlon 64 Desktop
- SDA – Sempron Desktop
3000 – processor rating:
- A - Socket 754
- D – Socket 939
- C-Socket 940
E – processor core voltage:
- E – 1.5 V
- I – 1.4 V
- A – 1.3 V
P – maximum processor core temperature:
- I – 63 °C
- P – 65 °C
- O – 69 °C
- K – 70 °C
- A – 71 °C
4 – second level cache size:
- 4 – 512 KB
- 5 – 1024 KB
- 6 – 2048 KB
AX – processor core type:
- AX, AW – Newcastle
- AP, AR, AS, AT – Clawhammer
- AK – Sledge Hammer
- BI – Winchester
- BN - San Diego
- BP, BW – Venice
- BV – Manchester
- CD – Toledo
- BI – Manchester (for Sempron)
- BA, BO, AW, BX, BP, BW – Palermo (for Sempron)
- AX – Paris (for Sempron)

For the second line, only four digits in the center are significant, corresponding to the week of processor production: 0529 - week 29 of 2005.

Intel processor markings

For processors from Intel, everything is a little more complicated - they only indicate the main characteristics, such as: clock frequency, processor number, system bus frequency, model marking like SL8FK. The rest of the information is provided only on the packaging of the boxed version. It is not easy to determine anything by the labeling of Intel processors, because there is no clear pattern in it. To find out the exact specifications, you need to go to processorfinder.intel.com and enter the markings there. Due to the fact that little is indicated on the processor itself, and in our country it is the versions without packaging (Tray or OEM) that are widely distributed due to their lower cost, then, unfortunately, there are no representative statistics regarding the release week of Intel processors, no matter how offensive it is . Therefore, in this article we will consider only processors from AMD, which are very popular today among domestic enthusiasts.

We talked about some characteristics of the Athlon 64 X2 series processors with two cores directly on the day of the announcement, but it was not possible to get a complete picture then. I would like to see official documents in which the missing characteristics would be clearly stated.

A little later we learned that processors with a 2 x 1 MB cache formula have a core area of 199 sq. mm, processors with a 2 x 512 KB cache formula have a core area of 147 sq. mm. In order to save chip area, the memory controller for both cores is made common. The first tests simulating the operation of the Athlon 64 4400+ (2.2 GHz) processor showed that this particular model has the optimal price-performance ratio. Only limited manufacturing capacity and a relatively large core area are preventing the Athlon 64 X2 from becoming widely available this year. AMD also partially admits this, talking about insignificant production volumes.

Since the announcement of Athlon 64 X2 processors based on the Toledo core will take place only in June, we would not be able to learn about the markings and characteristics of existing models soon, if not for one happy accident. One of our readers informed us that the special AMD Compare website now plays the role of a kind of analogue of the Intel Spec Finder, allowing you to find the necessary information about the characteristics of an AMD desktop processor in interactive mode. Documentation in PDF format on the official AMD website is now updated extremely rarely, so you can find out about the characteristics of new products at this address.

So, first of all, we determined the list of markings of existing Athlon 64 X2 models:

Athlon 64 4800+ (2.4 GHz, 2 x 1 MB) -> ADA4800DAA6CD(OEM), ADA4800CDBOX(BOX);
Athlon 64 4600+ (2.4 GHz, 2 x 512 KB) -> ADA4600DAA5BV(OEM), ADA4600BVBOX(BOX);
Athlon 64 4400+ (2.2 GHz, 2 x 1 MB) -> ADA4400DAA6CD(OEM), ADA4400CDBOX(BOX);
Athlon 64 4200+ (2.2 GHz, 2 x 512 KB) -> ADA4200DAA5BV(OEM), ADA4200BVBOX(BOX).

As we can see, models with 2 MB cache are based on E6 stepping, like Opteron processors with two cores, and models with 1 MB cache are based on E4 stepping, like San Diego processors. Moreover, even the part of the marking (“DAA”) that is responsible for the type of packaging, core voltage and maximum case temperature is the same for these processors. This means that all models of the Athlon 64 X2 series are Socket 939, operate at a core voltage of 1.35 V or 1.4 V, and the maximum case temperature is 65 degrees Celsius.

It is important that the official data says about the TDP level - it really is 110 W. As has been noted more than once, all motherboards compatible with the Athlon 64 FX-55 are suitable to support the Athlon 64 X2. After switching to Socket M2 and stepping F, Athlon 64 X2 processors will maintain the TDP level at 110 W, but the current will increase from 80 A to 95 A. Moreover, for processors of the Athlon 64 FX family, the TDP will increase from 104 W to 125 W. It cannot be ruled out that dual-core versions of the Athlon 64 FX will appear next year.

On the same site you can find a description of Sempron processors based on the Palermo core with E3 stepping. Finally, this stepping has a clear name. However, Palermo processors do not deny their relationship with the Venice core - at the current stage they are made from the same "raw materials", it's just that Sempron has part of the second level cache disabled.

It is characteristic that the E3 stepping processor has a lower maximum case temperature compared to its D0 stepping counterpart: 69 degrees versus 70 degrees Celsius. This may mean that the actual heat dissipation level of the Palermo-E0 and Venice cores is higher than their predecessors. Indeed, practical measurements have shown that the difference in power consumption reaches 16-17%. However, the frequency potential of the E stepping cores is much better, and for overclockers they are the best choice today.

Starting in 2011, Intel switched to the Intel Core labeling, which started with the second line. The currently used markings allow the user to quickly determine the required processor parameters.

Based on the labeling data for Intel processors, you can determine the connector for it, possible power consumption, and degree of cooling, because the more powerful the processor, the better the cooler should be.

Much also depends on the power supply, since processors with possible clock overclocking can consume much more energy than conventional ones. Therefore, the power supply must match the selected model.

Characteristics that determine the capabilities of the processor

The first parameter is the presence and number of cores in the chip itself: there can be two or four. Next, the number of threads is determined; Hyper-Threading technology is usually used, which controls the core threads. Equally important is the processor operating frequency, measured in gigahertz. This parameter is one of the few that can reflect the speed of the processor.

Starting with the i5 series, the manufacturer has introduced Turbo Boost technology, which allows you to increase the processor clock speed, which has a positive effect on performance. These are also the first processors to have four cores. Unfortunately, Intel Core i3 lacks these capabilities.

Another parameter is the cache; it is responsible for accelerated processing of data that is often used. The cache size ranges from 1 to 4 megabytes.

The last parameter determines the amount of heat removed from the processor to ensure normal operation of the CPU. The higher the processor temperature, the more powerful cooling is needed.

Step-by-step determination of the processor name

The first in the list of labeling for Intel Core processors is the name that the user pays most attention to. Next, the processor series is indicated and followed by a four-digit number, where the first digit is the generation, and the remaining three indicate the serial number. The last designation is a letter indicating the processor version.

For example, Intel Core i3 3200:

Intel Core is the name of the processor.
i3 means the third series.
3 - third generation.
200 is a serial number.

In this case, the Intel processor does not have a letter designation.

Characteristics of processor generations

In the marking of Intel processors, the first digit of the number indicates the generation, each of the digits corresponds to a specific name.

First on the list is the Westmere generation, which supports the DDR3 RAM format with frequencies of 1333 megahertz. There is no built-in video card. The technical process is 32 nanometers.

The next generation is called Sandy Bridge and supports RAM frequencies up to 1600 megahertz. The technical process is the same as in the previous version. The integrated graphics card is called Intel HD Graphics 3000.

The third generation is called Ivy Bridge and has a thinner technological process of 22 nanometers. RAM has not been changed. Intel HD Graphics 4000.

The fifth generation of Broadwell already works with DDR3L RAM (the letter prefix means a special connector) and frequencies up to 1600 megahertz. The manufacturing process is 14 nanometers thick, and the built-in graphics card is called Intel HD Graphics 6200.

The next generation, Skylake, had support for the DDR4 format and a 14-nanometer process technology. The built-in graphics component has acquired the three-digit designation Intel HD Graphics 580.

The latest known generation is Coffee Lake, which has completely switched to the DDR4 RAM format and a 14-nanometer process technology. The integrated graphics card is called Intel UHD Graphics 630.

Differences between processor series

The most common processor versions at the moment are i7. It is clear that the highest number means more powerful potential than a lower number. The i5 model is considered the most versatile option, since these processors can handle both basic tasks and complex applications.

Decoding letter indices

At the end of almost every Intel processor marking there is one letter, each of which carries a specific meaning.

H - designation for enhanced integrated graphics processor.
Q - from the word Quadro, means that the processor has four cores.
U - heat sink 15-17 Watt.
M - heat sink 35-37 Watt.
T - reducing the control of heat dissipation to 45 Watts.
S - lowering the control of dissipated heat to 65 Watts.
Y - lowering the control of dissipated heat to 11.5 Watts.
R - strengthening of the built-in video card for netbooks.
C - improved integrated graphics for LGA.
E - the presence of a chip with the function of embedding systems and a heat sink of up to 45 Watts.
P - disabled video core.
K is the overclocking potential of the processor.
X - presence of an Extreme chip.
M is a mobile processor, such a console belongs to the representatives of laptops.
MX is a mobile processor based on the Extreme chip.
MQ is a mobile processor with four cores.
HQ is a processor for laptops with support for high quality graphics.
L is a power efficient processor.
QE - the ability to integrate quad-core processors.
ME - embedded processors for laptops.
LE - presence of embedded processor optimization.
UE are processors whose optimization is aimed at optimal energy consumption.

Microprocessors from Intel

This type of processor has been known since 1971.

Microprocessors from this manufacturer can be 4-bit, 8-bit, 16-bit and 32-bit. The latest processors have proven themselves so well that they continued to be produced with the “Line” prefix. The differences between these processors are not only in the bus width, but also in the number of transistors.