by Ilya Gavrichenkov
12/04/2012 | 10:20 AM
Over the past few years AMD’s situation in the desktop processor market has been in decline. In the beginning the lack of progressive microarchitectures forced the company to slash the prices on their processors more and more, so that at some point they were completely gone from the top price segment featuring high-performance processors. After that they were in for a truly epic fail, when new processors on Bulldozer microarchitecture came out, which were the beacon of hope for the many. The Bulldozer was expected to become a product capable of competing against the top LGA 1155 and LGA 2011 Intel CPUs, but in reality the new microarchitecture turned out slow and power-hungry. As a result, Bulldozer became a niche product, capable of at least some resistance against the mainstream Intel processors only due to twice as many physical cores. Moreover, the performance could only be considered comparable during multi-threaded load and only if we disregarded the gigantic power consumption. In other words, Bulldozer launch didn’t really help AMD strengthen their positions in the desktop processor market.
Luckily, a chain of market failures combined with management issues, difficult financial situation, layoffs and search for new strategy didn’t interfere with the company engineers, and a year after Bulldozer launch we welcomed a second enhanced version of this microarchitecture called Piledriver. Our review of the top member of Vishera desktop processor family, FX-8350, showed that they didn’t waste any time over the past twelve months. FX-8350 managed to significantly improve the performance of AMD’s flagship platform. The benchmark results have increased by 15% on average, which is more than the performance boost Intel got upon transitioning from Sandy Bridge to Ivy Bridge microarchitecture. Trying to amplify the effect from their successful platform refresh, AMD chose to pursue very democratic pricing policy, so that Socket AM3+ platform became much more appealing than before. Vishera processors received a very warm welcome, and we also had absolutely no doubts that the new AMD offering would win a lot of users.
At the same time, if we dug a little deeper, we would see that processors with Piledriver microarchitecture are not that progressive after all. In fact, all the improvements are more of a superficial nature, and the advantage of the new FX-8350 over the FX-8150 can be explained by a combination of factors, most important ones being not the microarchitectural improvements, but the 400 MHz higher clock frequency and more aggressive Turbo mode. Of course, in the end the origin of the performance boost doesn’t matter that much for the end-users, but on the other hand, you should keep in mind that the advantage of the top Vishera processor model over the top Zambezi processor cannot be automatically projected onto other members of the corresponding families. AMD tried to push the clock frequency to the maximum in their FX-8350 processor, to make the best possible impression on the users, but simpler and less expensive FX processor modifications on the new microarchitecture, may fail to deliver a remarkable performance boost like that. Especially, taking into account that the new FX processor family also includes models with smaller L3 cache.
Therefore, we decided to investigate the performance of the junior Vishera processors, which were not part of our first review. Today we will try to determine if all FX processors on Piledriver microarchitecture can be considered a successful refresh of the family, or if only the new FX-8350 deserves the good words.
AMD sent out the top Vishera FX-8350 processor for the reviews of the new Socket AM3+ platform on Piledriver microarchitecture. However, in fact, the refreshed line-up includes four processor models: the above mentioned FX-8350, and slower modifications with lower clock frequencies featuring eight, six and four physical cores. They are all similar in some way: following the long-time tradition AMD sticks to unification within their product line-ups and uses a unified semiconductor dies in their Vishera processors. It consists of four dual-core modules and an 8 MB L3 cache. This semiconductor Vishera die consists of 1.2 billion transistors and is 315 sq.mm big.
However, the junior processors models do not use this entire die. The manufacturer may disable one or two dual-core modules, or cut off some of the L3 cache memory. On the one hand, this allows to easily create an entire line-up of diverse products, and on the other hand, the manufacturer gets the opportunity to put partially defective dies to good use. AMD varied the number of active dual-core modules in their previous generation Zambezi processors creating six-core FX-6000 and quad-core FX-4000 models. The differentiation in Vishera family has become much deeper: they also began to cut off half of the L3 cache memory. As a result, AMD FX processors family on the new Piledriver microarchitecture currently looks as follows:
Let’s take a closer look at the members of this family.
We have already discussed this flagship Vishera processor in an individual review. Nevertheless, I would like to remind you that it represents the most feature-rich modification of a Socket AM3+ processor available today. It means that FX-8350 has eight paired cores and 8 MB L3 cache. Its nominal clock frequency is set at 4.0 GHz, so the manufacturer has every right to claim that this is the first desktop processor to hit 4 GHz frequency level.
AMD believes that FX-8350 is a direct competitor to Intel Core i5-3570K, but in our opinion, this is an excessively optimistic prognosis. Nevertheless, the official price of the new FX processor is set at $195, thus allowing to compare it against slower Core i5 models on Ivy Bridge microarchitecture. Of course, the new AMD processor looks considerably more confident against those.
Unfortunately, we do not know yet how fast AMD is planning on increasing the clock speeds for their Vishera processors. It is very likely that FX-8350 will remain a flagship product for a while. Especially, since they have already postponed the launch of the next microarchitectural iteration codenamed Steamroller until 2014.
Obviously, not all Vishera semiconductor dies can work at 4.0 GHz frequency and 125 W max TDP, because they are manufactured using not the most advanced 32 nm process. Therefore, the top eight-core processor in AMD’s line-up is also accompanied by a slower modification – FX-8320 with similar characteristics, but lower clock speed. In other words, this processor, just like its elder brother, FX-8320, has eight paired cores and an 8 MB L3 cache, but works at 3.5 GHz nominal frequency, which is 500 MHz lower than that of the flagship processor. This processor can only speed up to the desired 4 GHz when you overclock it using Turbo Core technology. Nevertheless, the TDP of the slower Vishera processor still remains at 125 W.
Overall, the characteristics of the new FX-8320 are very similar to those of FX-8150 with Zambezi design. However, its price is set at $169, and it is positioned as a direct competitor to Intel Core i5-3450.
The six-core Vishera FX-6300 processor can be derived from a fully-functional semiconductor die by disabling one of the four dual-core processor modules. In terms of clock frequencies, this processor is similar to FX-8320. The nominal clock speed of this six-core processor is set at 3.5 GHz, but in case of fewer execution threads it may accelerate to 4.1 GHz. In this case the fewer cores and moderate clock frequencies allowed the manufacturer to set its TDP at less liberal 95 W. however, in the end, the working frequencies of the six-core Vishera processors yield to those of the FX-6200 based on the previous generation Bulldozer microarchitecture.
Disabling two cores in the original die didn’t affect the L3 cache memory size, which is also 8 MB big. However, the L2 cache in FX-6300 is smaller, because in Piledriver microarchitecture (just like in Bulldozer microarchitecture) it is individual for each pair of cores. As you may have guessed, it consists of three 2 MB parts, i.e. equals 6 MB against 8 MB in the eight-core processor.
Another distinguishing feature of the six-core Vishera modification is the lower frequency of the North Bridge integrated into the processor. This unit works at 3.2 GHz in FX-8350 and FX-8320, while in FX-6300 it is 200 MHz slower. It results into slight slowing of the processor memory controller and L3 cache. However, practical tests showed that this frequency difference was not critical.
However, FX-6300 is significantly more affordable than the eight-core processors. it is priced at $132 and is positioned as a direct competitor to the old Intel Core i5-2300.
FX-4300 seems to be the strangest FX processor from the new generation. It costs $122, which is only $10 cheaper than FX-6300, but at the same time it is significantly weaker in terms of technical characteristics. First, it has only four cores, because two dual-core modules out of four are disabled. Second, the L3 cache memory in FX-4300 has also been reduced to only 4 MB instead of 8 MB. As a result, we only get a “half” of the FX-8350, but not at half the price of the flagship product.
The operational frequencies of this processor are also far from impressive. The nominal speed of FX-4300 is 3.8 GHz, which could be accelerated to 4.0 GHz in Turbo mode. The North Bridge in this processor works at 2.0 GHz, just like the one in FX-6300. All this allows the quad-core Vishera to remain within the 95 W TDP, but at the same time it is significantly inferior to the previous generation top quad-core offering, the FX-4170, in technical specifications. The latter works at 200-300 MHz higher clock frequency and features a full-size L3 cache. Therefore, we have big concerns about the advantage of the quad-core Vishera processor over the previous generation quad-core Zambezi.
However, even AMD realizes the weakness of their FX-4300 specifications thus positioning their product against the dual-core Core i3-2120 from Sandy Bridge generation. Moreover, judging by the current prices, it would make more sense to go for a quad-core A10-5800K processor for Socket FM2 platform based on the same Piledriver microarchitecture. It sells at exactly the same price, but has an integrated graphics core and slightly higher clock frequencies.
Starting today we have significantly modified our testing methodology. Namely, we migrated to the latest Microsoft Windows 8 operating system. Of course, it doesn’t cause and radical changes, because the software environment doesn’t have too crucial of an effect on the computing performance of the tested platforms. However, it is important to keep in mind that Windows 8 kernel boasts a number of innovations.
First of all, unlike all previous operating systems, the scheduler in the new OS is originally optimized for all contemporary processor microarchitectures using SMT and CMT technologies. It means that neither contemporary AMD processors with paired cores, nor Intel processors with Hyper-Threading support will require patches of any kind: everything works in the most optimal way right out of the box. Secondly, the kernel of the new operating system did undergo some serious modifications improving its functioning on various mobile computers. And even though we focus on desktop tests, these changes matter for us, too. Windows 8 uses system memory more efficiently, parks the CPU cores more aggressively and tries to reduce the needs of the processor cycles. Thirdly, there is no more Aero interface in Windows 8: it has been replaced with Metro interface with higher 2D performance. And fourthly, the new system has been upgraded to DirectX version 11.1 (Direct3D 11.1, DXGI 1.2, WDDM 1.2, etc.).
As for the specific products tested today, AMD FX-8350, FX-8320, FX-6300 and FX-4300 from Vishera generation will compete against their predecessors on Zambezi design as well as against the contemporary competitors from Intel on Ivy Bridge. The honor of the old Bulldozer microarchitecture will be defended by the top processors in each category: eight-core FX-8150, six-core FX-6200 and quad-core FX-4170. Intel processors will be represented by the top LGA 1155 CPUs with four cores and Hyper-Threading support - Core i7-3770K, simpler quad-core products - Core i5-3570K and Core i5-3470, dual-core Core i3-3240 with Hyper-Threading support and the youngest member of the Ivy Bridge family - Pentium G2120.
As a result, our testbeds were built with the following hardware and software components:
As usual, we use Bapco SYSmark 2012 suite to estimate the processor performance in general-purpose tasks. It emulates the usage models in popular office and digital content creation and processing applications. The idea behind this test is fairly simple: it produces a single score characterizing the average computer performance. after the launch of Windows 8 SYSmark 2012 got updated to version 1.5, and this is exactly the version we are using in our test session.
Compared with the predecessors, Vishera processors make good progress. FX-8350 outperforms FX-8150 by 19%, which is mostly due to its higher clock frequency. However, you shouldn’t underestimate the contribution of the new microarchitecture, either. Even FX-8320 with slightly lower clock speed than that of the top Zambezi processor, is not any slower: FX-8320 is 10% faster than FX-8150. We see almost the same results among the six-core AMD processors from different generations: FX-6300 and FX-6200. However, the quad-core FX-4300 is only 6% ahead of the FX-4170, because the old quad-core Bulldozer processor works at a relatively high clock frequency, and also features a full-size 8 MB L3 cache, unlike its successor.
However, the success of the new AMD processors, which is so obvious when we compare the models from two different generations, doesn’t change the bigger picture involving competitors from Intel. Eight-core FX CPUs are still inferior to quad-core Ivy Bridge processors, while six-core and quad-core AMD processors can only compete against Core i3 or Pentium.
Let’s take a closer look at the performance scores SYSmark 2012 generates in different usage scenarios. Office Productivity scenario emulates typical office tasks, such as text editing, electronic tables processing, email and Internet surfing. This scenario uses the following applications: ABBYY FineReader Pro 10.0, Adobe Acrobat Pro 9, Adobe Flash Player 10.1, Microsoft Excel 2010, Microsoft Internet Explorer 9, Microsoft Outlook 2010, Microsoft PowerPoint 2010, Microsoft Word 2010 and WinZip Pro 14.5.
Media Creation scenario emulates the creation of a video clip using previously taken digital images and videos. Here they use popular Adobe suites: Photoshop CS5 Extended, Premiere Pro CS5 and After Effects CS5.
Web Development is a scenario emulating web-site designing. It uses the following applications: Adobe Photoshop CS5 Extended, Adobe Premiere Pro CS5, Adobe Dreamweaver CS5, Mozilla Firefox 3.6.8 and Microsoft Internet Explorer 9.
Data/Financial Analysis scenario is devoted to statistical analysis and prediction of market trends performed in Microsoft Excel 2010.
3D Modeling scenario is fully dedicated to 3D objects and rendering of static and dynamic scenes using Adobe Photoshop CS5 Extended, Autodesk 3ds Max 2011, Autodesk AutoCAD 2011 and Google SketchUp Pro 8.
The last scenario called System Management creates backups and installs software and updates. It involves several different versions of Mozilla Firefox Installer and WinZip Pro 14.5.
Although the total score may make Vishera processors seem not fast enough, there are popular situations when some of them do really well. Contemporary AMD microarchitectures are pretty weak when it comes to loads with few computing threads, because their performance-per-core turns out not high enough in the end. However, when we deal with multi-threaded loads, they can really shine because they have more physical cores than Intel offerings.
A few great examples of situations when the FX series processors do really well are 3D modeling tasks or financial calculations. Here eight-core FX-8350 and FX-8320 perform as good as the quad-core Core i5 processors and sometimes even better than them. However, the models with fewer cores can’t boast the same success. Even in the most favorable testing conditions, six-core and quad-core AMD Vishera processors can only get to the level of the Intel Core i3.
As you know, it is the graphics subsystem that determines the performance of the entire platform equipped with pretty high-speed processors in the majority of contemporary games. Therefore, we select the most CPU-dependent games and take the fps readings twice. The first test run is performed without antialiasing and in far not the highest screen resolutions. These settings allow us to determine how well the processors can cope with the gaming loads in general and how the tested CPUs will behave in the nearest future, when new faster graphics card models will be widely available. The second pass is performed with more real-life settings – in FullHD resolution and maximum FSAA settings. In our opinion, these results are less interesting, but they demonstrate clearly the level of performance we can expect from contemporary processors today.
Bulldozer microarchitecture didn’t do well in games. Luckily, its recent refresh, Piledriver, started showing signs of improvement in this aspect. Vishera has become significantly faster in games than Zambezi. As a result, FX-8150 is totally defeated not only by the new generation eight-core processors, FX-8350 and FX-8320, but also by the six-core FX-6300. However, the gaming performance of the new FX-4300 turned out quite disappointing. It is unfortunate that AMD decided to reduce its L3 cache memory size, because right now the new generation quad-core FX CPU loses even to its predecessor, FX-4170, in games that are sensitive to the memory sub-system performance.
However, speaking about the improvement of the gaming capabilities in the new Vishera processors with eight and six computing cores, it is important to remember that Intel CPUs continue dominating the gaming segment with much higher speeds. Core i7 and Core i5 based platforms produce more frames per second than systems with the top AMD FX processors, and Core i3 CPUs can easily challenge FX-6300.
It means that AMD fans enjoying occasional 3D gaming can only appeal to the fact that the actual gaming performance is limited by the graphics sub-system potential, which doesn’t let the CPUs fully shine. Therefore, in real-life situation, the difference between faster and slower processors may be hardly noticeable at all. However, it has to be a pretty weak argument after all. As we can see from the test results, there are games where the processor performance does affect the fps rate even with maximum image quality settings. Besides, there are new 3D shooters coming out these days, in which we know nothing about the effects of CPU performance on the graphics sub-system.
To test the processors performance during data archiving we resort to WinRAR archiving utility. Using maximum compression rate we archive a folder with multiple files with 1.1 GB total size.
WinRAR version 4.2 has been significantly optimized for multi-threading that is why AMD FX processors do pretty well here. This is exactly why eight-core FX from Vishera generation outperform Core i5, and the six-core FX-6300 starts breathing down its neck, too. However, the performance improvement we see in Piledriver products, doesn’t occur in FX-4300. AMD not only took away half of its cores, but also removed half f its L3 cache memory, setting it up to be defeated by FX-4170 and Core i3-3240.
The processor performance in cryptographic tasks is measured using a built-in benchmark of the popular TrueCrypt utility that uses AES-Twofish-Serpent “triple” encryption. I have to say that this utility not only loads any number of cores with work in a very efficient manner, but also supports special AES instructions.
Cryptographic tasks are a great way of uncovering the advantages of the AMD microarchitectures. Here FX-8350 works even faster than the top LGA 1155 Intel Core i7-3770K, and the FX-3820 with lower nominal clock speed falls just a tiny bit behind it. The six- and quad-core Piledriver processors also share the success of their flagship brothers. FX-6300 manages to outperform Core i5-3570K, while FX-4300 gets far ahead of the competitor’s Core i3 CPU. At the same time, new microarchitecture is not the primary contributing factor to this success. FX-8320 is only 2% faster than FX-8150, while FX-6300 is 5% ahead of FX-6200, and FX-4300 falls irreparably behind FX-4170. In other words, the 15% performance advantage of the flagship Vishera CPU over the corresponding Zambezi processor, which we admired during the very first test session featuring the newest member of the Piledriver FX family, should be attributed solely to the boosted clock frequencies.
We use Apple iTunes utility to test audio transcoding speed. During this test we transcode the contents of a CD disk into AAC format. Note that the peculiarity of this program is the exclusive single-threaded load.
Unfortunately, contemporary Socket AM3+ processors can’t offer us decent performance under single-threaded load. The weakness of Bulldozer (and now Piledriver) cores pushes even the top FX-8350 behind the Pentium G2120. Vishera processors have becomes a little faster than their predecessors, but it didn’t change anything in the global scope.
Now that the eighth version of the popular scientific Mathematica suite is available, we decided to bring it back as one of our regular benchmarks. We use MathematicaMark8 integrated into this suite to test the systems performance:
Wolfram Mathematica 8 is yet another example of an application where AMD products are in big trouble. We are not dealing with single-threaded load here, but the peculiarities of the FX processors microarchitecture again cause fiasco. Contemporary AMD processors only have one FPU per each dual-core module, which leads to low performance when working with floating-point numbers.
We measured the performance in Adobe Photoshop CS6 using our own benchmark made from Retouch Artists Photoshop Speed Test that has been creatively modified. It includes typical editing of four 24-megapixel images from a digital photo camera.
AMD FX processors can’t boast high performance results in Adobe Photoshop. Without the resource-hungry filters, but during the typical versatile image processing eight- and six-core FX processors on Piledriver microarchitecture turn out even slower than the dual-core Core i3-3240. However, this can still be considered an achievement, because the previous generation FX processors got defeated even by Pentium G2120.
We have also performed some tests in Adobe Photoshop Lightroom 4.2 program. The test scenario includes post-processing and export into JPEG format of two hundred 12-megapixel images in RAW format.
Adobe Lightroom knows how to process photographs using multiple threads, which immediately affects the results and improves AMD FX numbers. However, even though different Vishera models get as much as 16% faster than their Zambezi predecessors, they still cannot outperform Core i5 on Ivy Bridge. The junior FX from the new generation even falls behind FX-4170, which once again shows that cutting the L3 cache in half was a really bad idea for FX-4300.
The performance in Adobe Premiere Pro CS6 is determined by the time it takes to render a Blu-ray project with a HDV 1080p25 video into H.264 format and apply different special effects to it.
High definition video content processing is one of the best types of load for multi-core AMD processors. Besides, the new Piledriver microarchitecture works really well in applications like that. Vishera gets as much as 21% better than Zambezi in the same tests, which add up to the average of 12%. As a result, FX-8350 works better here than Core i5-3570K, FX-8320 almost reaches the level of Core i5-3470, and FX-6300 and FX-4300 are confidently ahead of Intel’s dual-core Ivy Bridge offerings, including the ones with Hyper-Threading support.
In order to measure how fast our testing participants can transcode a video into H.264 format we used x264 HD Benchmark 5.0. It works with an original MPEG-2 video recorded in 1080p resolution with 20 Mbps bitrate. I have to say that the results of this test are of great practical value, because the x264 codec is also part of numerous popular transcoding utilities, such as HandBrake, MeGUI, VirtualDub, etc.
HD video transcoding is another AMD-friendly task. We can clearly see it from the performance numbers demonstrated by the FX-8000 series processors. They prove capable of competing not only against Core i5, but even against the flagship LGA 1155 Core i7-3770K. However, high performance of the top Vishera processors during multi-threaded load is not typical of the simpler processor models. Note that FX-6300 is 70% behind FX-8350, while FX-4300 is even slower than half the speed of the eight-core CPU. As a result, the six- and quad-core Socket AM3+ processors can’t get even remotely close to the junior Core i5. They can only try to challenge Intel Core i3.
We will test computational performance and rendering speeds in Autodesk 3ds max 2011 using the special SPECapc for 3ds max 2011 benchmark:
Rendering is yet another example of a multi-threaded task, where AMD processors can unveil their strengths. New Piledriver microarchitecture does pretty well here. As a result, FX-8350 outperforms FX-8150 by as much as 20% and is even faster than Core i5-3570K. FX-8320 with lower clock speeds is only 6% faster than FX-8150, but it is more than enough for this processor to find its way into the pack of Core i5 Ivy Bridge CPUs. The six-core FX-6300 is 12% faster than the previous generation FX-6200, but it falls noticeably behind Core i5 series. As for the quad-core FX-4300, its performance is about the same as that of the FX-4170, which places it at the level of Intel Core i3.
When we talked about the FX-8350 processor, we arrived at the conclusion that Piledriver microarchitecture didn’t improve the energy-efficiency of the Socket AM3+ processors that much. AMD’s products manufactured with 32 nm process remain quite power-hungry compared with the 22 nm competitors. However, Vishera modifications reviewed today work at lower clock frequencies and have fewer cores, so they may turn out more energy-efficient than the flagship processor. Especially, since six- and quad-core modifications have 95 W maximum TDP instead of 125 W TDP of their “fully-fledged” brothers.
To find out more about the power consumption of all current processor models in the new AMD FX family, we performed a round of special tests. The new digital power supply unit from Corsair – AX1200i – allows monitoring consumed and produced electrical power, which we use actively during our power consumption tests. The graphs below (unless specified otherwise) show the full power draw of the computer (without the monitor) measured after the power supply. It is the total power consumption of all the system components. The PSU's efficiency is not taken into account. The CPUs are loaded by running the 64-bit version of LinX 0.6.4-AVX utility. Moreover, we enabled Turbo mode and all power-saving technologies to correctly measure computer's power draw in idle mode: C1E, C6, Enhanced Intel SpeedStep and AMD Cool’n’Quiet.
In idle mode all processors and platforms consume about the same power. In idle mode any contemporary processors will switch to special power-saving states, in which their power consumption will be very low – only a few watts. In this case the power appetites of other system components and the efficiency of the voltage regulator circuitry on the mainboard start to matter more and they partially mask the pure processor power consumption.
In case of single-threaded load things start going south for AMD processors. The fastest and most energy-hungry LGA 1155 Core i7-3770K consumes considerably less power than the most energy-efficient AMD FX product. At the same time, AMD processors work slower under this type of load than Intel CPUs. Yes, Vishera generation doesn’t waste the energy the way Zambezi processors used to, but there is still no visible indication of a qualitative change.
The situation doesn’t get any better when it comes to AMD FX power consumption under multi-threaded load either. FX-8350 turns out the most power-hungry CPU in our today’s test session. It consumes 4% more power than FX-8150 under maximum load and is 2/3 more power-demanding than Intel Core i7-3770K. Other members of the FX family on Vishera design do a little better compared with their predecessors, but they are still incomparable with any of the competitive offerings. At the same time, I have to say that FX-6300 and FX-4300 with 95 W TDP do look much batter against the background of all other Socket AM3+ processors. Had we also tested the old Core i5 processors manufactured with 32 nm process, then the new dual-module quad-core AMD processor would have been comparable against them in power consumption. Although in terms of performance this comparison makes absolutely no sense.
In other words, it looks like contemporary AMD processors are hopelessly behind their Intel competitors in terms of performance-per-watt. While in performance tests top eight-core Vishera processors can sometimes compete successfully against Intel’s quad-core Ivy Bridge CPUs, all illusions vanish into thin air the minute it gets down to power consumption.
AMD FX processor family is positioned as overclocker products. All these processors have unlocked frequency multipliers, i.e. support simple overclocking. Note that Zambezi and Vishera both have sufficient frequency potential for a substantial performance boost, which, however, will be accompanied by serious increase in power consumption. Nevertheless, overclocking is an important competitive advantage of the FX series CPUs. Intel processors can only be overclocked as easily only if they fall into the $220+ price range. As for AMD, they favor overclockers on a budget, which is one of the reasons Socket AM3+ platform has become increasingly popular.
Speaking of the Piledriver microarchitecture, AMD stressed that it would have increased frequency potential. Therefore, we were pretty optimistic about Vishera processors being more interesting to overclock than their predecessors, which could only go as far as 4.6 GHz on average with air-cooling. However, right from the start it became clear that the new processors didn’t improve that much after all – the first FX-8350 processor we received in our lab overclocked only to 4.7 GHz.
However, it would be unfair to judge the overclocking potential of an entire family just by one single processor, so we decided to check out the overclockability of all four Vishera products. The goal wasn’t to set overclocking records, but to determine the frequency, at which new FX processors would be able to continuously operate in 24/7 mode without losing stability. Therefore, we set the processor Vcore at 1.55 V, according to recommendations from AMD engineers. The cooling system used was NZXT Havik 140. After that we tested our overclocked system stability by running OCCT 4.3.2 utility (30-minute test in Large Data Set mode).
First of all, we retested our FX-8350 processor once again. We hoped to get a better result than the original 4.7 GHz. However, we have already checked out a few other FX-8350 samples since then and none of them did any better. So, it looks like 4.7 GHz is indeed a typical overclocking maximum for a flagship AMD FX processor equipped with an air cooler.
Many of you may think that the junior processor models are manufactured using not the best semiconductor dies that is why their overclocking potential is usually much lower. In fact, this is a pretty rare occurrence, and our FX-8320 unit proved that. It easily hit the 4.6 GHz mark, which is only 100 MHz lower than the results obtained on its more expensive counterpart.
The six-core FX-6300, which features one disabled dual-core module, overclocked practically as well as the “complete” Vishera processors. It remained perfectly stable at the maximum frequency of 4.7 GHz.
We were very optimistic about AMD FX-4300 overclocking, because some sources reported success in reaching 5 GHz mark with these CPUs, however, we couldn’t verify their reports. Our particular processor with half the active cores and only half the L3 cache remained stable at Vishera’s typical max frequency of 4.6 GHz. At least without the gambling at potentially dangerous Vcore levels.
So, it looks like all FX processors on Piledriver microarchitecture, independent of the number of cores they have, can overclock to about the same level of 4.6-4.7 GHz with air-cooling. Yes, it is better than what we saw from the previous generation of AMD FX processors, but there is certain no qualitative improvement in their frequency potential as of yet. Nevertheless, overclockers should be pretty happy with these results, which are quite typical of 32 nm processors.
I would also like to mention that when we overclocked FX-6300 and FX-4300 processors by raising their Vcore to 1.5-1.55 V, they didn’t heat up too much at all. The temperature of our six-core CPU rose to 65°C maximum, while the quad-core temperature stayed at the ridiculously low 53°C. It means that you could raise the core voltage higher and achieve stability at even higher clock speeds. However, we cannot approve of this approach, because excessive increase in the processor Vcore may lead to die degradation and therefore this mode is unacceptable for long-term use.
I have to say that the results of this test session have practically fully confirmed what we have already seen before in the first review of Socket AM3+ processors with Piledriver microarchitecture. The only difference is that this time we looked not only at the top CPU model, but at the entire product line-up. And this allowed us to somewhat revise our attitude to new AMD products. Here is why.
The flagship FX-8350 really does look very interesting. It is significantly faster than the previous generation AMD processors and can successfully compete against top LGA 1155 Ivy Bridge CPUs in case of multi-threaded load. Keeping in mind its affordable price, FX-8350 can be recommended for inexpensive desktops dealing primarily with such resource-demanding tasks as HD content creation and processing or final rendering. However, it is also important to keep in mind some of its drawbacks before you decide on this product. This processor is enormously power-hungry, and on top of that it is not universal as it doesn’t work fast in every-day general purpose tasks, which are mostly unable to split the load into eight parallel threads. I would also like to point out that 3D games are also among the problematic tasks for AMD processors.
Nevertheless, if you like FX-8350, then you should also consider FX-8320. This model is much cheaper, but offers practically the same level of performance – it will do great in professional applications. Moreover, since all contemporary Socket AM3+ processors belong to the Black Edition series, i.e. have unlocked clock frequency multipliers, FX-8320 can easily be overclocked to the level of the flagship CPU or even beyond that. This allows us to state that AMD FX-8320 is one of the most interesting choices for computer enthusiasts in terms of price-to-performance (multi-threaded). I wish it could eliminate the shortcomings that Vishera have: high power consumption and low performance in lightly-threaded applications. So, frankly speaking, FX-8320 is a good niche product, but not a general-purpose solution.
The six-core Vishera modification, FX-6300, seems to make a pretty weak overall impression at first glance. One of the four dual-core modules in this processor is disabled that is why its peak performance is relatively low compared with Intel’s quad-core processors even under multi-threaded load. This is quite logical, because two contemporary AMD cores are pretty much as fast as one Intel core: this is exactly what we saw throughout our today’s test session. However, the first impression is not always the right one, and it immediately disappears once you check out the price list. AMD priced their FX-6300 in such a way that it becomes a direct competitor to Core i3, instead of Core i5. And this strategic move offers Vishera new possibilities: they can be considered a really interesting alternative to Intel’s dual-core processors. Moreover, in this case FX-6300 even has a few trumps up its sleeve. For example, it can be overclocked unlike Core i3 processors.
However, the youngest memory of the new Vishera family, FX-4300 processor, turned out a total disappointment. Here AMD has obviously got carried away when chopping off its functionality: they not only disabled half of its cores, but also took away half of the L3 cache memory. As a result, FX-4300 is not any faster than FX-4170 with Bulldozer microarchitecture and is just a little more energy-efficient than the latter. So, the quad-core Vishera processor is noticeably slower than Intel Core i3, but at the same time it is priced very close to FX-6300. So, it looks like this particular processor model won’t be of any interest even to the loyal AMD fans, who will most likely prefer to go with a similarly priced A10-5800K with an integrated graphics core and higher clock frequencies.
That said, the appeal of AMD processors with Vishera design is not really in the advantages of the Piledriver microarchitecture, but in their low prices. In this respect, the two models in the middle of the line-up, FX-8320 and FX-6300, look best of all. These are the processors we would recommend checking out, if you are not discouraged by higher power bills. And please keep in mind that AMD processors perform best in multi-threaded tasks, but they are not as universal as Intel products. Therefore, Socket AM3+ platform probably won’t be a good choice for everyday use, and will best fit into an inexpensive workstation system.