AMD FX-8350 | Meet the Piledriver-based AMD FX processor line
As computer hardware reviewers, we are not very interested in the problems that hardware manufacturers face. Many would agree that the past year has been a terrible one for AMD's processor division, starting with the gluttonous Bulldozer CPUs, which slowly dropped in price in twelve months in response to the arrival of third-generation Intel Core processors. New samples have recently arrived at our laboratory in Southern California. End users shouldn't be too worried about talking about management flaws, layoffs and AMD's plagued past. So let's get down to business.
Sometimes it is possible to predict the ending of the article. If AMD sent us the FX-8170, which is 200 MHz faster than its predecessor, one would assume that the processor will show the same shortcomings in low-threaded applications, probably outperform the Core i5-2500K in more intensive tasks, but compared to the 77W chips. its power consumption will be just awful.
But instead we got a processor AMD FX-8350, which, by analogy with APU Trinity , introduced less than a month ago, is based on the Piledriver architecture. Experience tells us that Piledriver can outperform the Bulldozer design by 15% in performance per core and per clock cycle. It is also worth considering that AMD FX-8350 will run at least 400 MHz faster than FX-8150 ... By the way, the architecture provided a noticeably smaller increase in speed for the Intel Core line of processors. Chances are good that today's comparison will be much more interesting than a full blown FX-8150 in last year's review.
Meet the Piledriver FX Family
True to its tradition since Bulldozer, AMD has sent out to the press the fastest models in the new lineup, which will include eight, six and quad cores. Although they are all based on the Piledriver architecture, the chips themselves are called Vishera and will continue to be marketed under the FX brand.
AMD Vishera Crystal
The Vishera processor is 315 square millimeters and consists of 1.2 billion transistors. Exactly the same performance characterizes the previous generation Zambezi based on AMD Bulldozer architecture.
| 2012 AMD FX processor line | |||||||
| Number of cores / threads | Base frequency, GHz | Max Turbo Frequency, GHz | Northbridge frequency, MHz | TDP, W | Price, $ | OPN | |
| FX-8350 | 8/8 | 4 | 4,2 | 2200 | 125 | 195 | FD8350FRW8KHK |
| FX-8320 | 8/8 | 3,5 | 4 | 2200 | 125 | 169 | FD8320FRW8KHK |
| FX-6300 | 6/6 | 3,5 | 4,1 | 2000 | 95 | 132 | FD6300WMW6KHK |
| FX-4300 | 4/4 | 3,8 | 4 | 2000 | 95 | 122 | FD4300WMW4MHK |
Two of the four processors use eight processing cores, or four Piledriver modules, as AMD calls them. The base frequency of the flagship model AMD FX-8350 is 4 GHz. In low-threaded applications, Turbo Core technology can increase this figure to 4.2 GHz, although the increase in chip speed will be more related to the base frequency. How much Turbo Core accelerates AMD FX-8350? Not much. In the single-threaded iTunes benchmark, the result improved by only three seconds.
The FX-8320 lowers the base frequency to 3.5 GHz, but as part of the allocated thermal package Turbo Core increases it to 4 GHz (the 500 MHz boost is more important for the FX-8320). Both eight-core models are equipped with 8 MB L2 cache (divided into 2 MB for each module) and 8 MB L3 cache (shared for all four chip modules). Recommended price for AMD FX-8350 is $ 195, and the FX-8320 is $ 169.
The FX-6300 already has three active modules (six cores), and the price is reduced to $ 132. The 3.5 GHz clock gives the architecture an edge in multi-threaded tasks, and Turbo Core tries to compensate for the disadvantages in single-threaded applications by increasing the clock speed to 4.1 GHz. Like the four-module chips, the FX-6300 uses 2 MB of L2 cache per module (6 MB in total) and a shared L3 cache of 8 MB. Fewer active resources, as well as a lower frequency of the northbridge (2 GHz), allow the FX-6300 to be within the thermal package of 95 W, which is noticeably different from the 125 W of the older FX-83x0 processors.
The TDP of the dual-module FX-4300 processor also does not exceed 95 W. The base frequency of 3.8 GHz in low-bandwidth applications rises to 4 GHz, and the northbridge operates at 2 GHz like the FX-6300. However, the L3 cache capacity has been reduced to 4MB, and the price is only $ 10 lower than the three-module chip, prompting many buyers to spend an extra $ 10.
The AMD architecture does not feel much lack of memory bandwidth, however, the dual-channel DDR3 controller officially supports the data transfer rate of 1866 MT / s. To reduce the total cost of the platform, we will use DDR3-1600 modules with low latency, especially since the benchmarks (except for Sandra 2013 Beta) show that spending more money on faster memory will not gain speed.
The entire FX line has an unlocked multiplier, which makes overclocking much easier. Does Vishera have enough overclocking potential? What about the 5.125 GHz frequency supported by a closed loop liquid cooling system?
AMD FX-8350 | Overclocking and platform compatibility
Overclocking
Although AMD's processors have not been ranked the fastest for several years, the company is trying to keep the attention of enthusiasts by providing important features to them. Software that lets you change configuration settings in Windows in real time, unlocked multipliers and platforms with more PCI Express slots are just a few of the key features that AMD offers to people who know how to use them to increase the performance of their systems.
The same overclockers who were disappointed in the scalability of FX Zambezi processors with a conventional cooling system will be very surprised this time, even though we are considering a very similar architecture on the same 32nm core.
With CPU and Northbridge voltages of 1.375 V and 1.175 V, respectively, we managed to achieve stable operation AMD FX-8350 at 4.8 GHz at full load. The screenshot above shows a single-threaded test for spinning up the chip, but the highlighted maximum temperature corresponds to the peak load of the entire test package.
AMD FX-8350 could work even faster, but the key factor here is the voltage setting at which the temperature does not exceed 70 degrees Celsius. At this point, the temperature sensor forces the cores to underclock (the image above proves this), preventing the chip from overheating and negatively impacting performance. Without giving multi-threaded loads, we even managed to run tests at 5.125 GHz (for this, we need to set the voltage a to 1.4375 V for the CPU and 1.2 V for the northbridge).
Obviously, in most consumer systems, the bottleneck will be cooling. AMD's reference heatsink and fan are not enough, to put it mildly, and a powerful third-party solution will increase the cost of the platform with an FX processor. But solely for testing, we used the closed loop liquid cooling system that AMD offered with the FX processors last year. It will cost you about $ 70. If so, you can consider $ 300 as an alternative. Fortunately, we have test results.
Overclocking to 4.8 GHz is enough to AMD FX-8350 overtook in a multi-threaded environment such as 3ds Max 2012, however it did not help AMD Piledriver architecture get around in a single threaded iTunes application. Of course, if you are willing to spend an extra $ 30 on and even more on a third-party cooler, then its base frequency can be relatively easily increased from 3.5 GHz to 4.5 GHz and will be measured against AMD FX overclocking.
Compatibility
All measurements were made using a Mastech MY64 multimeter.
Search for software to detect instability
The software chosen to detect instability can be roughly divided into three categories:
- Programs originally targeted as stress tests of the system. This category includes LinX 0.6.4 (testing was carried out in 2560 MB mode for the old version of Linpack, as well as in three modes, with available memory of 1024 MB, 2560 MB and 6144 MB for the latest version of Linpack, with support for FMA instructions), OCCT 4.3.2.b01 (CPU test: OCCT in Large Data Set, Medium Data Set and Small Data Set modes, and CPU: LINPACK test in AVX mode with 90% of available memory), Prime95 v27.7 build2 (in Small FFTs, In-place Large FFTs and Blend modes), CST 0.20.01a (a combined test that includes the Matrix \u003d 5, Matrix \u003d 7 and Matrix \u003d 15 modes).
- Programs that are used as tests of system performance, or that emulate one or another load that occurs in everyday PC work. Got here Cinebench R10 (x CPU test), Cinebench R11.5 (CPU test), wPrime 1.55 (test 1024M), POV-Ray v3.7 RC3 (All CPU's test), TOC [email protected] Bench v.0.4.8.1 (Dgromacs 2 test), 3DMark 06 (CPU1 + CPU2 test), 3DMark Vantage (CPU1 + CPU2 test) and 3DMark 11 (this time, separately Physics Test and separately Combined Test).
- Several processor-dependent games. These included Colin mcrae dirt 2 Deus Ex: Human Revolution (Detroit), F1-2010 (built-in performance test), Metro 2033 (built-in performance test), Shogun 2 Total War (Battle of Okehadzam) and The Elder Scrolls V: Skyrim (The estate "Golden Flower").
For stability, the state of the system is taken, in which no problems arise in its operation within 10-15 minutes of the test.
CPU instability
In this subsection of the article, we will choose software that makes it easier to identify the instability of the processor, with obviously stable memory and CPU_NB frequencies. The technique is relatively simple: with a fixed value of the supply voltage, select the maximum overclocking for each of the programs and calculate the test at which the minimum frequency of stable operation will be reached. Well, in parallel with the search for stable frequencies, you can also evaluate the behavior of the system during overclocking for a particular test. To avoid instability caused by overheating of the CPU, all tests were performed with the CPU supply voltage of 1.25 V.
advertising
The processor frequency at which Windows starts is 4256 MHz.
Yes, this is not a joke of humor, and a new model is announced based on the well-worn AMD 990FX chipset with an SB950 south bridge.
The solution looks a little odd. So, for example, the north bridge could be renamed 1000FX or 1100FX, and not even anyone would not notice the catch. But with the south bridge, everything is not so rosy. It's still a little outdated. Just a little because there is no support for the USB 3.0 interface. Therefore, it was necessary to add four ports of this standard, plus a couple more SATA 3 and call it all SB1050. But this did not happen, and a platform for enthusiasts based on the old logic set is being announced.
advertising
Is he really that bad? Probably not, because there is support for basic functions and interfaces. Yes, PCI-e is of the second generation, but there is support for both SLI and CrossFire modes. In addition, due to the age of the chipset, one can already expect that prices for products based on it will be reduced. And the owners of motherboards based on it can sleep peacefully and continue to be on top, since it remains an advanced solution.
And if support for USB 3.0 can be added with third-party controllers, then with PCI-e 3.0 things are worse. There is one more thing - this is the integrated video core. Historically, the FX series processors have to compete with the Intel s1155 platform. On it, most processors come with an integrated video adapter. You can argue for a long time about the fact that it is not needed, but it is given for free. It allows you to not only watch movies and play casual games, but sometimes it can help you with your work. Therefore, considering all these aspects, AMD needs to be more serious about its successor. Perhaps because of this, the announcement had to be postponed.
Translation of the official instructions from ASUS ROG.
As you know, overclocking your processor instance may differ from the results of this article. The final result can also be influenced by the cooling and the motherboard (if you are using an unspecified board).
If you have additional questions, you can ask them on the official forum (English) or in the official community on the social network Vkontakte.
Well, let's get started!
Installing a processor in a socket is quite simple and you cannot use any effort, otherwise you can damage the legs, and, accordingly, kill it:
-lift the clamp
-install the processor by aligning the golden triangle with the triulet mark on the socket
-low down the clamp
Apply a thin layer of thermal paste. It is very convenient to use something like a credit card if there was no special spatula in the kit. 
Install cooling. Remember that overclocking the FX-8350 requires very good cooling! We use the Corsair H80i in our tutorial, but analogs are fine. It is best to use a two-piece CBO.
Installing RAM. 
A little about the system power supply.
As you may have already noticed, ASUS ROG Crosshair V Formula-Z is equipped with an additional 4-pin processor power connector. Connecting both cables (8-pin and 4-pin) can improve overclocking stability. But if your power supply only has one 8-pin cable, then don't worry - connecting the second is not necessary at all. 
Before we go directly to overclocking, download the latest BIOS for your board and unpack it to a flash drive so that you can update the board via EZ Flash. 
You also need to download (or take on the disk from the kit) and install the ASUS AI Suite II utility, since not all utilities can correctly monitor the temperatures of the FX-8350. We'll be using AI Suite II to monitor temperatures and voltages, CPU-z to monitor CPU frequency, and Prime95 to check system stability.
Make sure everything is working as expected.
Go to BIOS and press F5 to apply optimal settings.
Save the settings with the F10 key and reboot into Windows.
Open CPU-Z and AI Suite II.
Launch Prime95 and run Small FFT test. The processor frequency should be approximately 4 GHz.
In our case, the processor temperature reached 49 degrees. The room temperature was 25 degrees. 
If everything is in order, then let's go directly to overclocking!
Note again that there can be a very large difference between processors, but most should be overclocked to 4.8 GHz with good cooling. If your system is unstable, then try lowering the multiplier one step at the same voltage and check it again.
We will use the extended UEFI BIOS mode (ROG boards initially run this mode).
Follow the Extreme Tweaker tab.
Let's set the CPU Ratio parameter to 24. Thus, we should get 4.8 GHz (you can see the Target CPU Speed \u200b\u200bvalue a little higher). 
Set the memory mode to your preference (according to your memory specification). In our case, 1600 MHz. 
In the same tab below, look for CPU & NB Voltage and set it to Manual mode.
Set the CPU Manual Voltage to 1.5 V. Ignore the red backlighting of the numbers, everything is in order. 
Go up the menu a little higher and go to DIGI + Power Control.
Set CPU Load Line Calibration to Ultra High. 
Now we need the Advanced tab. In it we go to the CPU Configuration menu. 
The Cool ‘n’ Quiet must be set to Always Disabled and C1E to Disabled. 
You can also disable (Disabled) Core C6 State.
You can save the settings profile (so as not to set everything again each time) in the Tool tab.
Now press the F10 key to save the settings and reboot.
Stress test
I will repeat everything as last time. Launch AI Suite II, CPU-z and then Prime95 in Small FFT mode.
In our case, the temperature in the load increased to 76 degrees. That's quite a lot, but still within the TJMax range. 
Is the system stable? Then go back to BIOS and try to lower the voltage to 1.49 V and below. You can try changing the Load line Calibration value to High, but most likely the system will not work stably. Our processor unit was able to run at 4.8 GHz at 1.49 V. 
If you have additional questions, then you can ask them
