One of the most productive microprocessor devices within the AM3 + socket was the AMD FX-8350 chip . The characteristics of this solution, its real level of performance, reviews about it and other important information will be further given with respect to such a CPU within this material. Comparative testing will also be conducted with competing semiconductor products.
CPU device positioning
CPUs were divided as follows in the framework of the AM3 + computer platform:
The most budget PCs were based on the chips of the Septron and Athlon series. They had very modest parameters, minimum operating frequencies and a truncated cache. But the price tag for these decisions was appropriate.
Entry-level system units were based on the FX-43XX series. This series of processors without fail included 4 computing modules, had a 3-level “fast” memory and an unlocked multiplier. But the price in this case increased significantly.
Mid-level gaming systems, in turn, were based on FX-63XX chips. Their only difference from the previous CPU series is the presence of 6 modules for processing code instead of 4.
Premium PCs were based on microprocessors of the FX-9XXX and FX-83XX series. These devices already included 8 clusters of code and data processing. It was to this line of semiconductor solutions that the AMD FX-8350 considered in this review belonged . His characteristics were really the best within this platform.
The processor socket AM3 + debuted in 2012 and is currently morally and physically outdated. He was replaced in 2017 by a more advanced socket under the code designation AM4. It is he who provides the greatest performance among AMD solutions.
Types of trim levels. What is included in them?
In two possible configurations, you can find the AMD FX-8350 processor. The characteristics of one of them indicate the absence of a cooling system. In this case, thermal grease and cooler are purchased separately. It is indicated on the Trail price lists. For the rest, the manufacturer included in the delivery list the following:
BOX is the second possible configuration of this microprocessor device. In it, the manufacturer included a standard cooling system from a proprietary cooler and thermal paste, this made it possible to use the characteristics of the AMD FX-8350 in nominal mode . A review of the model indicates an unlocked multiplier. Therefore, for overclocking, it is recommended to buy the first delivery option in combination with an advanced and improved heat sink system. If the processor is planned to be used in nominal mode, then the VOH option is out of competition in this case.
Chip architecture
The presence of 8 computing units focused on the characteristics of the AMD FX-8350. The photo of the semiconductor crystal and the given structural diagrams of its organization indicate a not quite standard layout of this solution. The number of processing units for integers is really 8 pieces. But there are only 4 modules for floating-point processing on this silicon base. As a result, it can be noted that this chip is equipped with only 4 paired cores, each of which has 2 integer devices and one floating point. Such an organization significantly reduces the speed that the AMD FX-8350 processor boasts. Its performance characteristics indicate that it functions faster than the i3 series, but slower than the i5. The devices of the first series can process 8 instructions per cycle (2 cores of 4 instructions each), and the second - 16 (4 cores of 4 instructions each). The hero of this article can also process 16 instructions (8 cores, 2 teams each). But the program code is optimized specifically for Intel chips, as a result, they show better results in tests. But all this is true for the nominal mode. If you overclock the FX-8350, then it will be able to compete on equal terms with the i5 4th or even 5th generation. The code name for the microprocessor architecture of this solution is Vishera. And this CPU was produced using 32nm technology.
Clock frequencies
The frequency formula in AMD FX-8350 is rather unusual. Its technical characteristics indicate the basic value of this parameter at 4.0 GHz. But also in this CPU, TurboCore technology is implemented, which automatically increases this value to 4.4 GHz. In addition to this, the multiplier of this chip is unlocked. As a result, it can be dispersed with virtually no significant problems. But for this, as noted earlier, the cooling system must have an improved design. In this case, the CPU can be overclocked to 5.0-5.2 GHz. These values apply to air heat dissipation systems. If we use liquid nitrogen-based solutions for this, then these chips can overcome 5.5 GHz.
RAM memory
The outdated layout of the organization of the volatile memory subsystem is indicated by the characteristics of the AMD FX-8350. Owner reviews at the same time indicate a decrease in the performance of the computing system for this reason. Unlike modern PCs, the RAM controller is moved outside the CPU crystal and is included in the chipset of the motherboard. It can address 32 GB of memory (4 modules with a nominal size of 8 GB) and is able to function in 2-channel mode. Type of chips - DDR3. Their clock frequency can be 1066, 1333, 1600 or even 1866 MHz. You can use a faster solution, but their frequency will be no more than the maximum allowable 1866 MHz. But this is true for the nominal mode. If you overclock the PC, then the frequency of the modules of the RAM subsystem can exceed this limit.
Cache
A fairly good cache organization is indicated by the AMD FX-8350 review . The characteristics of this microprocessor focus on its three-level organization. The most difficult is the first one. It includes 4 clusters of 64 kb each, which can only store processor instructions, and 8 blocks of 16 kb for data. In total, this allows you to store 192 kb. The second level includes 4 clusters of 2 MB. In this case, strict specialization regarding the type of stored memory in this case does not exist. Its total size is 8 MB. The third level of the very “fast” volatile memory in the AMD FX-8350 is most simply organized . Its characteristics indicate the presence of 8 MB. And in this case, it consists of only one segment. Again, there is no distinction regarding the type of information stored in this case.
Energy consumption. Temperature
A very high power consumption is indicated by the AMD FX-8350 review. The characteristics of the processor indicate the presence of a thermal package of 125 watts. This is very important today. Its modern analogues have a thermal package of not more than 95 watts. The high energy consumption of the hero of this review is due to the outdated technological process for the production of a silicon crystal, which corresponds to 32 nm. Current chips are manufactured with tolerances of 14 nm. The temperature regime of this chip should not exceed 61 ° C. In the nominal mode, this value ranges from 50 to 57 ° C. After overclocking with a standard cooling system, problems may occur with overheating of the microprocessor and the temperature may exceed the maximum allowable temperature of 61 ° C. Therefore in case of overclocking, the computer system should be equipped with an improved cooling system. If this condition is met, then you can count on the acceleration of such a CPU to a frequency of 5.0-5.2 GHz and the temperature will not exceed 57 ° C.
Microprocessor overclocking
As has been noted many times before, the AMD FX-8350 boasts a high level of overclocking potential . The characteristics of the WRAITH cooler, which comes with the VOX configuration, allow this chip to be overclocked to 4.5-4.6 GHz. Again, the case must provide improved air circulation and be equipped with several additional fans. If you purchase the Trail CPU delivery option in combination with an improved fan and advanced thermal grease, then you can expect to achieve 5.2 GHz. The most optimal solution for acceleration is liquid nitrogen-based cooling. Such systems are much more expensive, but you can already count on the fact that the limit of 5.5-5.6 GHz will be reached. In this form, this microprocessor comes close to the results of the 4th generation Intel i7 CPU.
Performance
Now let's test AMD FXTM - 8350. The characteristics of this chip indicate that it can compete on equal terms with the microprocessors of the i3 and i5 series. As a result, as the opponent on the part of Intel it is most optimal to choose the solution of the last series with the index 3570K. For comparison, on the AMD side, it is most optimal to choose the FX-8150 and Phenom II with an index of 1100T. These are the main competitors of the microprocessor considered in this material. In the PCMark7 synthetic test, these chips received such points:
3570K - 6873.
FX-8350 - 6539.
FX-8150 - 6057.
1100T - 5402.
This test is optimized for single-threaded execution, and due to this feature, the Intel solution bypasses any AMD chip. In turn, AMD microprocessors are distributed quite standardly in order to improve the microarchitecture that underlies them. In the 3DMark11 test suite, the alignment of forces in synthetic points undergoes certain changes:
FX-8350 - 8502.
3570K - 8082.
FX-8150 - 7623.
1100T - 7082.
AMD’s flagship bypasses the middle Intel in this test suite, and there is a simple explanation for this - the software is optimized for multi-threaded execution and the improved architecture of the blue company in this case loses all its advantages. When processing multimedia information or archiving, AMD chips look better, but in games, Intel products are preferable. An example is the results in conditional points in the WinRAR archiver:
FX-8350 - 7112.
FX-8150 - 6268.
3570K - 5571.
1100T - 5470.
Multithreaded optimization in this case makes itself felt, and the Intel chip lags far behind the two AMD flagships. Now we give the tests in the game F1 2012 with a resolution of 1920x1080 and high quality pictures in FPS:
3570K - 97.
FX-8350 - 81.
FX-8150 - 67.
1100T - 63.
All CPUs provide an acceptable level of fps, but the more frames per minute, the more comfortable the gameplay. Therefore, the average from Intel in this case is beyond competition. A similar arrangement of chips in the game RUSE The number of fps in a similar mode is obtained in this way as follows:
3570K - 47.
FX-8350 - 39.
FX-8150 - 37.
1100T - 26.
The advantage of i5 from Intel is already 22 percent over the hero of this article. Again, most of the games are optimized for single-threaded execution and the decisions of the “blue” company in this case are simply better and preferable.
Cost
Fairly acceptable specifications for the AMD FX-8350. Owner reviews this feature of this microchip only confirm. But at the same time, his cost is quite democratic. To date, the Trail delivery option can be purchased for 7,000 rubles. But Voh costs a little more expensive - about 8000 rubles. In the first case, it is necessary to add the cost of the cooling system to the cost of the chip. If you purchase an advanced air cooling system, you need to add 3,000 rubles. When using liquid cooling, this amount will have to be increased to 5,000 rubles. As a result, it should be noted that the VOH option is most suitable for working in the nominal operating mode, and the second is better for overclocking the CPU.
Reviews
Although this chip was released back in 2012, it still allows you to solve any problems and the characteristics of the AMD FX-8350 are still relevant . Reviews about the model really confirm this. Of course, he does not reach the level of the most productive computers. But he still allows you to create medium-performance computing systems. It is most optimally used for processing software that is optimized for multi-threaded implementation. This is transcoding video or audio files. Also in this group belong graphic editors. Also this microprocessor has a quite affordable price. Its only drawback is the increased energy consumption, which, as previously noted, is due to production technology. Alas, somehow getting rid of it in this case simply will not work.
Conclusion
A rather mixed impression is the AMD FX-8350. Its characteristics are quite relevant, but it is highly discouraged to consider it as the basis for the assembly of a new high-performance computer. The solutions of the AM3 + series are currently outdated, they are gradually being replaced by AM4 chips, which debuted at the beginning of 2017. Therefore, the scope of the chip considered in this material is farms for mining various kinds of cryptocurrencies or server computers. In this case, the requirements for productivity and energy efficiency are minimized, and the cost of the final system comes to the fore. Any microprocessor within the AM3 + socket, including the FX-8350, can boast of such a combination. In other cases, it is better to purchase Ryzen series CPUs for the AM4 connector. After all, but this semiconductor solution is still outdated. Also, Intel chips can be considered as an alternative. And in this case, it is more preferable to choose the solutions of the LGA1151 series. They have performance and, of course, performance is an order of magnitude better than that of the FX-8350. In addition to this, their energy efficiency is also an order of magnitude better, which is ensured by an improved production technology of 14 nm. Any Ryzen series microprocessor can boast the same feature. The only plus of the device in question against their background is the low cost. Of course, the cost of 7000-8000 rubles for the flagship processor solution is very, very democratic. If you need to assemble a high-performance PC with minimal investment, then such a chip is very suitable. But at the same time, it is absolutely necessary to take into account the moment that further modernization of such a computing system will be impossible. The AM3 + socket is outdated today, and AM4 and LGA1151 are gradually replacing it. The only advantage of the legacy AM3 + platform is its lower cost. For the rest, these chips have completely and completely lost their relevance for the reason that the CPU capabilities in the most labor-intensive modes are no longer enough.