| AMD FX-8150 Bulldozer Processor |
| Reviews - Featured Reviews: Processors | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Written by David Ramsey | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Wednesday, 12 October 2011 | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
AMD FX-8150 Bulldozer CPU ReviewThe days when AMD processors ruled the performance roost are long gone, and most enthusiasts have forgotten they ever existed. But less than a decade ago, the Athlon X2 dual-core processor thoroughly spanked Intel's first crude dual-core efforts, which were simply two separate processor dies on one chip, communicating through the front-side bus, as opposed to AMD's much more advanced true dual-core CPUs. But AMD's been playing catch-up since then, and has been forced to compete on price rather than performance in desktop processors. Now Benchmark Reviews tests the high end of the new Bulldozer desktop CPU line, the multiplier-unlocked 8-core FX-8150 CPU. Enthusiasts have been slow to adopt multi-core CPUs beyond dual cores. Few software used them (especially games), and while extra cores could help a system running multiple applications, only a few individual applications could really leverage the extra power provided by multiple cores. But multi-core processor penetration in the market is increasing: while the largest percentage of users in the Steam hardware survey use only dual-core processors (47.6% as of August, 2011), four-core CPUs are close behind at 43.5%. Beyond four cores, the numbers drop off dramatically. Only 1.45% are using six-core processors, and 0.07% have 8 cores (and most of these are probably dual four-core processors). So the FX-8150 has an open market...
I've been a fan of AMD processors for some time, unbothered by their performance deficit relative to Intel since we've long since passed the point where almost any processor is more than fast enough for anything most people would want to do, and AMD simply offered better bang for the buck. But AMD's got a tougher challenge this time in Intel's new Sandy Bridge processors, not to mention the upcoming Sandy Bridge E series and Ivy Bridge. AMD already cedes the high end to Intel when they say the FX-8150 is designed to compete against the Core i7 2500K CPU rather than the top-end 2600K, but even that will be a strong competitor: it's even less expensive, at $219.99 at Newegg compared to the FX-8150's MSRP of $245.00.
Manufacturer: Advanced Micro Devices Full Disclosure: The product sample used in this article has been provided by AMD. AMD FX-8150 FeaturesThe following information is courtesy of AMD
AMD FX CPU SpecificationsAMD's consumer CPU line now has three main divisions: the E series APUs, which are aimed at netbooks, small notebooks, and tablets; the A series APUs for desktops and higher-end laptops (Benchmark Reviews tested the AMD A8-3850 "Lynx" CPU in this review), and the top-of-the-line FX processors. Unlike Intel's Sandy Bridge CPUs, all of which have integrated video, the FX CPUs are just CPUs. AMD expects you to add a Radeon 6000 series card to an FX system. The following chart shows the specifications of all the desktop-level FX-series processors. At launch, only the FX-8150, FX-8120, FX-6100 and FX-4100 will be available.
AMD's initial lineup of FX CPUs includes three eight-core chips, one six-core chip, and three four-core chips. Aside from core count and clock speed, the chip specifications are similar, with the main difference being that the amount of L2 cache is lower for the six and four-core CPUs. All FX CPUs are unlocked, so you'll be able to tweak clock speeds to the limits your hardware can handle. A surprisingly high DDR3-1866 memory speed should help bandwidth, assuming you have memory that can run at that speed. AMD's Turbo Core technology will use the maximum speed if only half or fewer of the cores are in use; if all cores are in use, Turbo Core is dialed back to keep the CPU within its thermal limits. With the new AM3+ socket come three new chipsets: the 990FX, the 990X, and the 970X, all supported by the SB950 south bridge. All the new chipsets provide six SATA 6G ports, 12 USB 2.0 ports, and RAID levels 0, 1, 5, and 10. The only difference is in the supported PCI-E configuration: 2x16 or 4x8 for the 990FX, 2x8 for the 990X, and 1x16 for the 970X. Benchmark Reviews covered the 990FX chipset in our review of the ASUS Crosshair V Formula motherboard. Obviously, the exciting thing about the FX-8150 is that it's the first consumer eight core processor. Prior to the AMD FX, if you wanted an eight-core CPU you had to buy a very expensive server-level Xeon or Opteron chip, and then you'd have to deal with things like the special "registered" memory these CPUs require, and an expensive server motherboard lacking many of the features expected in enthusiast-level motherboards...like, say, overclocking. (Yes, I know about the Intel Skulltrail and EVGA SR2 motherboards. But they're the exception.) But will eight cores really make a difference? Well, that's what we're here to find out. In the next section I'll go into the details of the Bulldozer processor architecture. Bulldozer CPU ArchitectureTechnically, the term "Bulldozer" refers to the two-core module that AMD uses as a building block for its new processors. Desktop CPUs built with Bulldozer modules are code-named "Zambezi", while server processors built with Bulldozer modules are "Interlagos" for CPUs designed for single and dual processor systems, and "Valencia" for processors designed for four or more CPU systems. Here's a block diagram of a Bulldozer module: 
Each module comprises two integer execution units and one floating point execution unit. All three units share the instruction fetch and decoders, as well as the L2 cache, but each core has its own instruction scheduler and L1 cache. Each module can handle two concurrent threads (one on each integer unit), so AMD considers a single Bulldozer module to be "dual core." But is this an accurate description? Intel supplies this labeled die image for Sandy Bridge processors: 
Note that there are four areas labeled "Core". For comparison, here's a labeled die image for a quad-module (8 cores) AMD Bulldozer CPU: 
Here, there are four areas labeled "Bulldozer module", and as we saw above, each module has two cores. But what does a "core" consist of? In Sandy Bridge, each core contains three Arithmetic Logic Units (ALUs) and two Address Generation Units (AGUs). AMD's Phenom II CPU has three ALUs and three AGUs per core, while a Bulldozer "integer core" has two ALUs and two AGUs, for a total of four ALUs and four AGUs per module. (The floating point execution unit is a separate entity. Since floating point instructions comprise only a small percentage of most code, the single FPU in a Bulldozer module is shared.) An Arithmetic Logic Unit does the actual work of handling the instruction, be it a conditional, bit rotate, add, or other integer operation. The Address Generator Unit handles address generation, and to explain that would involve getting deep into Intel address architecture, which is beyond the scope of this article. Suffice it to say that the AGU is needed for everything from figuring out the real address of that branch destination to where to put the results of a given calculation to translating between virtual memory addresses and physical addresses. Keeping these units "fed" are the instruction schedulers, which decide how to dispatch instructions that have been fetched from memory and decoded. A Phenom II core can issue 3 ALU or AGU instructions per clock; Sandy Bridge can issue 3 ALU/1 AGU or 2 ALU/2 AGU instructions per clock (four total), while Bulldozer can issue 2 ALU/2 AGU per clock (also for a total of four). Note that while the Phenom II core has the most ALUs/AGUs (6 total), it can issue the fewest instructions per clock (three, as opposed to four for Sandy Bridge and Bulldozer). There are other differences: each Sandy Bridge core has its own floating point unit, while a Bulldozer module has two integer units (the ALUs) and only a single floating point unit. So the eight core FX-8150 has four floating point units, just like the four-core Sandy Bridge. Since the vast majority of program instructions will be handled by the integer ALUs, this shouldn't affect real-world performance...but our benchmarks will show if this is the case. Complicating comparisons further are things like decode queues, instruction pipelines, branch prediction, thread retirement, cache management, and more. Modern processors are hideously complex; long gone are the days when a programmer could figure out exactly how long a segment of code would take to execute by simply adding up the number of clock cycles required for each instruction and figuring the time based on the CPU clock frequency! (Yes, we really used to do that.) Bulldozer adds further optimizations: for example, if only one integer core is being used, it has access to all the resources on the module, such as the cache: there's nothing "reserved" for the idle core. AMD has also finally caught up with some of Intel's new features, like AES-NI for ultra-fast encryption and decryption, Advanced Vector Instructions (AVX), and a 32nm fabrication process (which will hopefully help with overclocking). AMD also has some new instructions of their own: the FMA (floating point multiply-accumulate) and XOP instructions. The former allows fast multiply-and-add sequences, which turn out to be useful in video transcoding, among other things. The XOP instructions are an extension to SSE5 that AMD announced back in 2009 and consist of new integer vector instructions, such as integer vector multiply-accumulate and integer vector compare; the Bulldozer CPUs are the first implementation of XOP. All of these features sound really cool when AMD's technical and marketing folks are giving the presentations. But unless you're a CPU architect (and I'm not), it's impossible to judge their real-world impact on paper. And while I'm convinced some of AMD's new internals are very advanced, Intel's made some enormous strides with Sandy Bridge, especially in the crucial "instructions per clock" metric. Simply put, this means that at the same frequency as previous generation Intel CPUs, Sandy Bridge is significantly faster. And of course many Intel processors support Hyper-Threading, which doubles the number of available cores (as far as software is concerned), so that a four-core processor appears as an eight-core processor. AMD claims that eight real cores should provide better performance than four real and four virtual cores; while this makes sense, AMD is positioning the FX-8150, the top of the Zambezi desktop processor lineup, against the non-HyperThreading 2500K rather than the Hyper-Threading capable Core i7-2600K. Well, all this makes for interesting arguments in bars, but what really matters is the performance. So let's get on with the testing. Processor Testing MethodologyFor this test Benchmark Reviews is fortunate to have two high-end 990FX motherboards: AMD supplied an ASUS Crosshair V Formula with a pre-installed FX-8150, and MSI supplied their new 990FXA-GD80. Benchmark Reviews examined the ASUS Crosshair V Formula motherboard here, and a detailed review of the MSI 990FXA-GD80 will follow shortly. Although I tested the AMD FX-8150 in both motherboards, all the overclocking and performance measurements for this article were done on the ASUS motherboard. For comparison purposes I also used an AMD Phenom II X6-1100T six core CPU, AMD's previous top of the line part. Representing the Intel camp is what AMD considers the FX-8150's competition: a Core i5 2500K CPU, in this case installed in an ASUS P8Z68V Pro motherboard. Although I used the same memory for both Intel and AMD tests (a G. Skill 4GB kit rated at up to 2133MHz), I set the memory speed to the maximum officially supported by each CPU/chipset: 1333MHz in the case of the Intel system and the 1100T, and 1866MHz for the AMD FX-8150 systems. The memory timings were 9-9-9-24 in all cases. Different overclockers have different techniques. For example, AMD's recent 8.4gHz world record overclock of the FX 8150 represented the maximum speed at which they could successfully boot into Windows and run CPU-Z. Other overclockers go for the maximum single-core overclock. I prefer to aim for the highest speed at which I can run all cores under load. The FX 8150, like the Intel Sandy Bridge CPUs, uses a 32nm fabrication process, and reaped similar overclocking benefits: on the ASUS Crosshair V Formula motherboard, I was able to hit 4.8gHz on all cores at 1.4V and complete all my benchmarks. This helped performance significantly, but the Core i5 2500K can be overclocked, too...and coincidentally my test 2500K hit its highest stable overclock at the same 4.8GHz frequency as the FX-8150! The overclocked results for both processors are included in all the benchmark tests. Intel Z68 Test Platform
AMD 990FX Test Platforms
Benchmark Applications
AIDA64 Extreme Edition TestsAIDA64 Extreme Edition is the evolution of Lavalys' "Everest Ultimate Edition". Hungarian developer FinalWire acquired the rights to Everest in late November 2010, and renamed the product "AIDA64". The Everest product was discontinued and FinalWire is offering 1-year license keys to those with active Everest keys. AIDA64 is a full 64-bit benchmark and test suite utilizing MMX, 3DNow! and SSE instruction set extensions, and will scale up to 32 processor cores. An enhanced 64-bit System Stability Test module is also available to stress the whole system to its limits. For legacy processors all benchmarks and the System Stability Test are available in 32-bit versions as well. Additionally, AIDA64 adds new hardware to its database, including 300 solid-state drives. On top of the usual ATA auto-detect information the new SSD database enables AIDA64 to display flash memory type, controller model, physical dimensions, and data transfer performance data. AIDA64 v1.00 also implements SSD-specific SMART disk health information for Indilinx, Intel, JMicron, Samsung, and SandForce controllers. All of the benchmarks used in this test— Memory reads and writes, Queen, Photoworxx, ZLib, hash, and AES— rely on basic x86 instructions, and consume very little system memory while also being aware of Hyper-Threading, multi-processors, and multi-core processors. Of all the tests in this review, AIDA64 is the one that best isolates the processor's performance from the rest of the system. While this is useful in that it more directly compares processor performance, readers should remember that virtually no "real world" programs will mirror these results. Although I'm using the same physical memory on both the AMD and Intel systems, I'm running it at different speeds: the officially supported maximum on each platform, which is 1333MHz for the Intel Core i5 and AMD Phenom II 1100T, and DDR3-1866 for the FX-8150. Let's see how this plays out in AIDA64's memory throughput tests: 
And...not that well. While the FX-8150 shows a huge improvement in memory throughput over its predecessor, the Core I5 2500K still easily wins, even doubling the write throughput performance. This is an interesting result considering the higher frequency of the FX-8150 memory subsystem. 
The Queen and Photoworxx tests are synthetic benchmarks that iterate the function many times and over-exaggerate what the real-world performance would be like. The Queen benchmark focuses on the branch prediction capabilities and misprediction penalties of the CPU. It does this by finding possible solutions to the classic queen problem on a chessboard. At the same clock speed theoretically the processor with the shorter pipeline and smaller misprediction penalties will attain higher benchmark scores. Like the Queen benchmark, the Photoworxx tests for penalties against pipeline architecture. The synthetic Photoworxx benchmark stresses the integer arithmetic and multiplication execution units of the CPU and also the memory subsystem. Due to the fact that this test performs high memory read/write traffic, it cannot effectively scale in situations where more than two processing threads are used, so quad-core processors with Hyper-Threading have no real advantage. The AIDIA64 Photoworxx benchmark performs the following tasks on a very large RGB image:
The Intel processor produces the highest scores in both benchmarks, but the differences are interesting: at stock clocks speeds, the 2500K and FX-8150 are neck and neck, with the 1100T actually very slightly in the lead. Overclocking the Intel results in a decisive win, with the Intel CPU coming in about 18% above the AMD. In the Photoworxx benchmark, the FX-8150 is much faster than the 1100t, even faster than its two extra cores would lead you to believe. The Intel processor still wins, although not by as large a margin as it did in the Queen benchmark, and the very high overclocks for both processors produce only minor improvements in their scores. 
In the ZLib test, the AMD CPUs leap ahead of the Intel 2500K, with even the previous-generation 1100T beating the stock-clocked 2500K. In the Hash test, the difference is even more profound: the Intel CPU simply can't keep up with the AMD CPUs here. Intel's Clarksdale and subsequent CPUs have dominated the AES test due to their Advanced Encryption Standard New Instructions (AES-NI), which dramatically accelerate AES code. AMD's own implementation of AES-NI makes its first appearance in Bulldozer-based CPUs, and in the ASUS motherboard turns in slightly better scores than the Intel CPU, although the score in the MSI motherboard is, oddly, about 11% lower. Overclocking has a minimal effect on this benchmark in either case, and the Phenom II X6-1100T, which does without AES-NI, just can't compete. 
So far, we've seen the Intel and AMD CPUs slug it out and swap wins in these tests. Let's move on to the PCMark Vantage benchmark. PCMark Vantage TestsPCMark Vantage is an objective hardware performance benchmark tool for PCs running 32- and 64-bit versions of Microsoft Windows Vista or Windows 7. It's well suited for benchmarking any type of Microsoft Windows Vista/7 PC: from multimedia home entertainment systems and laptops, to dedicated workstations and high-end gaming rigs. Benchmark Reviews has decided to use a few select tests from the suite to simulate real-world processor usage in this article. Our tests were conducted on 64-bit Windows 7, with results displayed in the chart below. TV and Movies Suite
Gaming Suite*
Music Suite
* EDITOR'S NOTE: Hopefully our readers will carefully consider how relevant PCMark Vantage is as a "real-world" benchmark, since many of the tests rely on unrelated hardware components. For example, per the FutureMark PCMark Vantage White Paper document, Gaming test #2 weighs the storage device for 100% of the test score. In fact, according to PCMark Vantage the video card only impacts 23% of the total gaming score, but the CPU represents 37% of the final score. As our tests in this article (and many others) have already proven, gaming performance has a lot more to do with the GPU than the CPU, and especially more than the hard drive or SSD (which is worth 38% of the final gaming performance score). 
The TV and Movies suite concentrates on video playback and transcoding, but only uses two threads at a maximum, so the extra cores of the AMD processors shouldn't be an advantage. The FX-8150 is less than 4% slower than the 2500K, but overclocking the Intel CPU adds enough performance to significantly increase this lead, while overclocking the FX-8150 leads to a relatively small performance gain. The Gaming benchmark relies on the hard disk and video card for over 50% of its score (see the Editor's Note above), and we're using the same HDD and video card for all platforms, so the Intel processor's decisive win in this test simply means that Vantage's gaming code is more optimized for Intel processors. Bear in mind, however, that most "real world" games will not show this difference; generally, in games, your video card matters most, followed by the clock speed (not number of cores) of your processor. The PCMark Vantage gaming test can use up to 16 threads, so it's very strange that the eight core FX-8150 turns in markedly lower scores than the six core 1100T. Unlike the Gaming test, the Music test results have more real-world relevance, since multi-threading is much more common in music transcoding applications than it is in games. Even so, the four core Intel CPU beats the six and eight core AMD processors. The wins are piling up in the Intel column. Let's move on to CINEBENCH. CINEBENCH R11.5 BenchmarksMaxon CINEBENCH is a real-world test suite that assesses the computer's performance capabilities. CINEBENCH is based on Maxon's award-winning animation software, Cinema 4D, which is used extensively by studios and production houses worldwide for 3D content creation. Maxon software has been used in blockbuster movies such as Spider-Man, Star Wars, The Chronicles of Narnia, and many more. CINEBENCH Release 11.5 includes the ability to more accurately test the industry's latest hardware, including systems with up to 64 processor threads, and the testing environment better reflects the expectations of today's production demands. A more streamlined interface makes testing systems and reading results incredibly straightforward. The CINEBENCH R11.5 test scenario comprises three tests: an OpenGL-based test that models a simple car chase, and single-core and multi-core versions of a CPU-bound computation using all of a system's processing power to render a photo-realistic 3D scene, "No Keyframes", the viral animation by AixSponza. This scene makes use of various algorithms to stress all available processor cores, and all rendering is performed by the CPU: the graphics card is not involved except as a display device. The multi-core version of the rendering benchmark uses as many cores as the processor has, including the "virtual cores" in processors that support Hyper-Threading. The resulting "CineMark" is a dimensionless number only useful for comparisons with results generated from the same version of CINEBENCH. For this test I used only the multi-core rendering benchmark. All AMD processors beat the 2500K at stock clock speeds, but overclocking leads to another win for Intel, although by a mere 7%.
Looking at the results, we see a four core processor holding its own or beating six and eight core processors. The obvious conclusion is that Intel gets more done "per core" than does AMD. Later in this review I'll take a closer look at single core performance. For now, it's on to CPU dependent 3D gaming.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Base | Single Core Max | L1 Cache | L2 Cache | L3 Cache | DDR3 Freq | |
| FX-8150 | 3.6GHz | 4.2GHz | 128K | 2M (shared per module) | 2M (shared per module) | 1866MHz |
| Phenom II 1100T | 3.3GHz | 3.7GHz | 128K | 512K | 6M (shared) | 1333Mhz |
| Core i5 2500K | 3.3GHz | 3.7GHz | 64K | 256K | 6M (shared) | 1333Mhz |
Comparisons between differing core architectures are inherently imprecise. Recall that Bulldozer modules each comprise two integer cores and one shared floating point core, and that if only a single integer core (APU) is in use, it can use all of the normally shared cache resources. For these tests I used benchmarks that had "number of threads" settings and set the number of threads to "1". First up is the CINEBENCH single core rendering test.
Intel dominates here. While the FX-8150 could leverage its eight core to roughly match the 2500K in multi-core rendering, it's no contest at the single core level.
\
And Intel's domination continues in single-core renderings using Blender and POV-Ray. Let's look at Passmark:
AMD leads in three of these four tests, including, oddly, the integer test. Intel traditionally beats AMD in integer performance, so I can't explain this result, especially given the very large 70% delta.
But the FX-8150 falls behind again in three of the last four Passmark tests.
Note that in most of these tests, the spiffy new Bulldozer cores in the FX-8150 provide about the same performance as the Thuban cores in the Phenom II X6-1100T. This is disappointing especially given the base and turbo frequency advantage enjoyed by the new processor. But AMD has one more trick up their sleeve: new x86 instructions that applications can use to increase performance.
AMD FX New Instructions
The basic "x86 instruction set" has been with us since the eight-bit Intel 8080 processor was introduced in 1974. Since then, both Intel and AMD have introduced extensions to the x86 instructions like MMX, SSE, 3D Now, and so forth. In the Bulldozer, AMD has added these instructions:
- AES New Instructions-- Originally introduced by Intel, these instructions dramatically accelerate AES encryption.
- AVX-- Advanced Vector Extensions enhance floating point performance.
- FMA4 and XOP-- New instructions for complex math, useful in rendering and video transcoding applications.
The problem with new instructions is that existing programs won't benefit, since the instructions didn't exist at the time the programs were built. New instructions require support from compiler vendors, who must upgrade their products to generate the new code when appropriate (or directed by the programmer). Current benchmarks and applications will not benefit from these new instructions, but new ones will.
To demonstrate the performance improvements possible with these new instructions, AMD provided recompiled versions of x264HD using the new AVX and XOP instruction features. Both the 2500K and the FX-8150 support AVX instructions, but the 1100T does not.
For the first two runs, Intel wins, although if you compare the FX-8150's Run 1 scores against its scores in the standard x264HD benchmark, you'll see a 61% improvement, from 75fps to 120.61fps. Intel improves too, from 93.42 to 150.7fps, which is also a 61% improvement. In Run 2, the performance deltas are similar.
AMD pulls ahead in Runs 3 and 4, but the improvements in frame rates over the standard x264 benchmark are minimal. Now, let's look at the results from x264HD coded with the XOP instructions.
Since the XOP instructions are currently unique to the Bulldozer, the Sandy Bridge processor and the 1100T cannot run this benchmark at all. However, the scores returned for the FX-8150 are virtually identical to the ones returned by the AVX version of the benchmark in the first two runs.
The pattern repeats itself in Runs 3 and 4. The scores for the FX-8150 are about the same as they were for the AVX version of the benchmark.
AMD says that current Microsoft compilers incorporate the switches to enable code generation for these instructions. Based on this single benchmark, they can provide significantly improved performance that dramatically narrows Intel's performance lead.
AMD FX-8150 Overclocking
Like Intel's Sandy Bridge, AMD's new Bulldozer processors are fabricated on a 32nm process. This means they use less power and generate less heat than the older Phenom II CPUs, which were based on a 45nm process. Less power and less heat generally means better overclocking, and the FX-8150 doesn't disappoint: I was able to reach a stable 4.8GHz on all eight cores under extended stress testing, running the CPU at 1.4V and using a Thermalright Silver Arrow air cooler. This is 200MH higher than AMD suggested was feasible for all cores "on air" in their reviewer's guide, so I feel pretty good about it! With the giant Thermalright Silver Arrow cooler, processor temperatures under load maxed out at 59 degrees at an ambient temperature of 24 degrees.
Overclockers were initially dismayed that Intel's Sandy Bridge and its supporting Cougar Point chipset removed two classic overclocking mechanisms: increasing the base clock speed and increasing the base multiplier. You can only overclock Sandy Bridge CPUs by increasing the turbo multiplier, and then only on a "K" series Sandy Bridge CPU and a supported motherboard chipset. Somewhat ameliorating these limitations was the fact that if you did have the right hardware, Sandy Bridge CPUs were capable of immense overclocks.
AMD's FX series CPUs, in contrast, are completely unlocked and the 990FX chipset does not generate the base clock for the entire system, so you have all the classic overclocking mechanisms available. Now, for most real world use, overclocking by raising the turbo multiplier is your best bet: the CPU will ramp up its clock speed under load as needed, but drop back to a lower clock speed when it's idling, using less power and generating less heat. But for my tests I wanted to be sure that the Bulldozer was pushing with its blade all the way down under all circumstances, so I raised the base multiplier to 24, which when multiplied by the 200MHz base clock resulted in a speed of 4.8GHz. I did try a 24.5 multiplier, and I could boot and run at 4.9 but the system would crash under stress testing.
The chart below shows how overclocking affected the 8150's performance on each benchmark, with the stock clocked benchmark score normalized to 1.0 and the overclocked benchmark score represented as how much faster it was than the base clocked benchmark.
| 8150 Stock | 8150 OC | |
| AIDA64 Memory Read | 1.0 | 1.02 |
|
AIDA64 Memory Write |
1.0 | 1.05 |
| AIDA64 Queen | 1.0 | 1.20 |
| AIDA64 Photoworxx | 1.0 | 1.05 |
| AIDA64 AES | 1.0 | 1.00 |
|
AIDA64 ZLIB |
1.0 | 1.14 |
| AIDA64 Hash | 1.0 | 1.14 |
| CINEBENCH Single Core | 1.0 | 1.13 |
| CINEBENCH Multi Core | 1.0 | 1.19 |
| Passmark CPU Marks | 1.0 | 1.24 |
| PCMark Vantage TV/Movies | 1.0 | 1.05 |
| PCMArk Vantage Gaming | 1.0 | 1.11 |
| PCMark Vantage Music | 1.0 | 1.04 |
| Street Fighter IV | 1.0 | 1.12 |
| Handbrake 0.95 | 1.0 | 1.16 |
| x264HD Run 1 | 1.0 | 1.21 |
| x264HD Run 2 | 1.0 | 1.19 |
| x264HD Run 3 | 1.0 | 1.18 |
| x264HD Run 4 | 1.0 | 1.17 |
| SPECviewperf Lightwave | 1.0 | 1.13 |
| SPECviewperf Maya | 1.0 | 1.10 |
| SPECviewperf TCVIS | 1.0 | 1.16 |
| SPECapc Lightwave Interactive | 1.0 | 1.09 |
| SPECapc Lightwave Multitasking | 1.0 | 1.15 |
| SPECapc Lightwave Rendering | 1.0 | 1.20 |
| Blender/Icetest | 1.0 | 1.24 |
| POV-Ray | 1.0 | 1.13 |
| Average Performance Improvement | 13.3% | |
The performance improvement is less than what I'd hoped, but bear in mind that the AMD FX-8150 can crank its cores up to 4.2GHz at stock clock speeds, and the difference between 4.2GHz and 4.8GHz is only 14.9%, so the observed 13.3 percent is quite close to what you'd expect from the clock speed difference.
Late note: AMD will be selling the FX-8150 in a kit bundled with a water cooler. Although there isn't an MSRP yet, AMD says the price will be "about $100" more than the CPU by itself. Benchmark Reviews received the water cooler too late to incorporate into this review, but we'll be investigating its performance in a future article.
Bulldozer Final Thoughts
Those of us hoping that AMD's Bulldozer CPUs would catapult ahead of Sandy Bridge performance must live with the disappointment. Considered on its own, the AMD FX-8150 CPU is a great processor with excellent performance (especially if you can keep all eight cores busy), and in many cases beats the Intel Core i5 2500K. But there are few real-world applications that gamers and enthusiasts use that will fully exploit it; in fact, as you can see from many of these benchmarks, even programs designed to spawn multiple threads frequently do not scale their performance well past four cores.
Of course, having eight cores also means you can have a lot of background stuff going on and still keep things perky in the foreground application, and AMD has some persuasive demo videos showing how much smoother multitasking is with real cores as opposed to virtual cores, but it's a hard thing to quantify.
The 32nm fabrication process and other architecture improvements give the FX-8150 good overclocking headroom, and AMD's aggressive turbo frequencies represent the first officially-supported 4GHz and higher clock speed in any consumer CPU. But the Core i5 2500K has lots of overclocking headroom, too. Since overclocking results aren't guaranteed, the comparison chart below is based on each CPU's scores at stock clock speeds.
| Stock clocked performance results | Core i5 2500K score | FX-8150 score | FX-8150 % Diff |
| CINEBENCH single | 1.5 | 1.01 | -33% |
| CINEBENCH multi | 5.45 | 5.93 | +8.8% |
| SPECviewperf Maya | 14.86 | 9.71 | -14.9% |
| SPECviewperf TCVIS | 7.44 | 5.05 | -32% |
| SPECviewperf LightWave | 31.83 | 26.43 | -17% |
| SPECapc multitasking (lower is better) | 1389 | 1351 | -2.7% |
| SPECapc interactive (lower is better) | 535 | 674 | +26% |
| SPECapc rendering (lower is better) | 871 | 792 | -9% |
| AIDA64 Memory Read | 16508 | 13838 | -16% |
| AIDA64 Memory Write | 18341 | 10255 | -44% |
| AIDA64 Queen | 32278 | 31702 | -1.8% |
| AIDA64 PhotoWorxx | 52548 | 46598 | -11.3% |
| AIDA64 Zip | 198 | 262 | +32.3% |
| AIDA64 AES | 375440 | 380271 | +1.3% |
| AIDA64 Hash | 2071 | 3677 | +77.5% |
| Vantage TV & Movies | 5988 | 5753 | -3.9% |
| Vantage Gaming | 11787 | 6173 | -47.6% |
| Vantage Music | 11017 | 9901 | -10.1% |
| Street Fighter IV low-res (fps) | 568 | 410 | -27.8% |
| Handbrake (times in secs, lower is better) | 146 | 127 | +15% |
| x264Bench HD Run 1 | 93 | 75 | -19.4% |
| x264Bench HD Run 2 | 97 | 75 | -22.7% |
| x264Bench HD Run 3 | 29 | 36 | +24.1% |
| x264Bench HD Run 4 | 29 | 36 | +24.1% |
| PassMark CPUMark | 6987 | 8657 | +23.9 |
| Blender/icetest rendering time (lower is better) | 112 | 118 | -5.1% |
| POV-ray rendering time (lower is better) | 269.29 | 214.06 | +25.8% |
Using these benchmarks, the stock-clocked AMD FX-8150 averages 2.2% slower than the stock-clocked Intel Core i5 2500K. But as you can see looking at this chart, the individual differences are typically much higher: the FX-8150 tends to win big or lose big. It would be easy to choose a mix of benchmarks that gave a decisive win to either CPU, but I tried to use as broad an array of tests as I could to give the most accurate performance comparison.
As I showed in the single-core section, the performance of a Bulldozer core is not significantly better than the performance of the older AMD Thuban core, and both are far behind a Sandy Bridge core, so AMD's banking on keeping all eight cores filled to get the best performance. And indeed the FX-8150 can return excellent performance in these cases, although the performance improvement is less than what you might expect given the extra cores. And if software vendors upgrade their products to use the new instructions AMD has integrated into Bulldozer, its performance will improve more.
AMD claims the Windows 7 thread scheduler doesn't make the best use of Bulldozer's architecture, and says that we can expect a 10-15% performance improvement when Windows 8 ships. Also, Bulldozer is just the first in a line of new processors: in the coming years we'll see Piledriver (2012), Steamroller (2013) and Excavator (2014), each of which AMD says will bring improvements in performance-per-watt and instructions-per-clock.
But Intel's not standing still. Before the end of the year we should see Sandy Bridge "E" processors, and next year we'll see the 22nm Ivy Bridge processors, which according to rumor will drop right into existing Cougar Point motherboards, enabling Intel users to easily upgrade their systems. AMD FX CPUs are only officially supported on AMD 9-series chipsets, although several vendors have said that their 8-series motherboards will support FX processors with a BIOS upgrade. When specifcally asked about this, AMD said only that official support is limited to the 9-series chipsets, but that individual vendors were free to do what they wanted. We'll see, I suppose.
Given the market for this processor, you might wonder why I didn't include game benchmarks. The reason is simple: at high resolutions (say 1080p and above) and with multi-monitor systems, the critical factor is your video card setup, not the processor. Assuming the same video card configuration, your Bulldozer gaming rig will not give you a noticeably different experience than your Sandy Bridge gaming rig. If you're an AMD fanboy like me, rest assured that a Bulldozer system is an excellent base on which to build a high end gaming rig.
A few weeks ago I attended an all-day press briefing on the Bulldozer architecture at AMD's Austin, Texas headquarters. One of the points AMD made was that performance is only one aspect of a processor; another is performance per watt. But while AMD's server processors and E- and A-series APUs excel in this area, the FX-8150's 125 watt TDP is 32% greater than the Core i5 2500K's 95 watts. Granted, most desktop system users don't consider processor power draw when designing a system, but there it is.
AMD's weakness remains its integer core (APU) performance, which Bulldozer does not significantly improve over Thuban. More cores can compensate for this in some circumstances, but overall in my tests the FX-8150 can be considered at best to have achieved performance parity with Intel's 2500K...at a $254 MSRP as compared to the 2500K's $216 MSRP.
AMD FX-8150 Conclusion
Benchmark tests should always be taken with a grain of salt. It's difficult to try and isolate the performance difference a single component in a computer system makes, especially when it's necessary to compare across different manufacturers and platforms. Complicating the matter is the fact that benchmarks change, a manufacturer may change the technical details of a product, and the retail price may change as well. So please use this review as just one piece of information, and do your research before making a buying decision.
AMD fans have been awaiting the Bulldozer for months, and while it represents a significant performance improvement over Thuban, I'm disappointed with its overall performance, especially the memory bandwidth and single-core performance. Prior to Sandy Bridge, the FX-8150 would have been a very competitive processor; now, the best you can say is that performance-wise it's as good, overall, as a Core i5 2500K. Except that the Intel processor costs about $35 less: and that's a sore point, because historically AMD processors tend to beat Intel in a "bang for the buck" competition due to their lower prices.
As of November 2011, the previous-generation "Thuban" 1100T six-core processor sells for a mere $190 while Bulldozer costs $269.99 at Newegg, so AMD fans will want to carefully consider if two more cores are worth another $64 and a new motherboard, especially given that the per-core performance is virtually the same between the Bulldozer and Thuban cores.
We can hope that Windows 8 and upgraded applications and utilities that use the new FX instructions will make it more competitive, and I'd expect these things right about the time Ivy Bridge become available.
Pros:
+ First consumer eight-core processor
+ Officially supports 4GHz-plus turbo speeds and DDR3-1866 memory
+ An FX system has 42 PCI-E lanes as opposed to the 24 lanes of a Sandy Bridge system
+ 990FX chipset supports NVIDIA SLI. Finally.
+ AMD finally has a 32nm processor with good overclocking
Cons:
- Requires a new Socket AM3+ motherboard- Single core performance has remained static
- Full performance requires Windows 8 system and applications that use its new instructions
- Overall similar performance to Core i5 2500K, but at a higher price
Ratings:
-
Performance: 8.00
-
Construction: 9.00
-
Overclock: 9.50
-
Functionality: 8.50
-
Value: 8.00
Final Score: 8.60 out of 10.
Benchmark Reviews invites you to leave constructive feedback below, or ask questions in our Discussion Forum.
Related Articles:
- Antec Notebook Cooler 200
- Palit GeForce 9800 GTX+ Video Card NE/98TX+XT352
- Seagate FreeAgent Go 640GB External Hard Drive
- QNAP TS-259 Pro Turbo NAS Server
- Patriot Inferno SSD Kit PI100GS25SSDR
- Thermaltake MEKA G-Unit Gaming Keyboard
- G.SKILL 2133MHz DDR3 Pi-Series Memory Kit
- NETGEAR R6300 Gigabit Wi-Fi Router
- QNAP TS-809 Pro 8-Bay SATA NAS
- Gigabyte GA-EX58-UD4P X58 Motherboard
Comments
BTW, why would you bet on a chip?
Every time someone releases a product there are people like you who dump all over it as if you've been personally harmed. Please get some perspective.
The perspective that needs to be gained here is that he has as much right to his opinion as you do to yours without being labeled as "dumping all over it". You work for AMD or something?
ASUS M5A87 Mobo
AMD Bulldozer FX 8120 BE
16GB DDR3 RAM 1600MHZ
Radeon HD 6950 2GB GDDR5 (1408 Cores)
Cougar CMX 1000W PSU
(unfortunately running on stock cooling)
so far i've been really impressed with the chip battlefield 3 in % cpu usage barely tickles the 8120 you could run the game twice and still have enough left over for antivirus defrag or god knows what else. And yes i am a little disappointed with its gaming bench tests i read in PC Format. But still its doing me really really well.
My "Intel system" is an overclocked Core i7-3960X CPU and two GTX580 video cards in SLI. My "AMD system" is an overclocked FX-8150 and a single Radeon 6950 video card.
With a 27", 1920x1200 monitor, I can't see any difference in any of the games I play between one system and the other. Sure, I can measure frame rate differences with benchmarks, but these aren't noticeable in actual game play.
The situation would likely be different were I running, say, a triple-monitor setup. But how many people actually do?
Fanboy alert...
#goldsborowebdevelop ment.com/wp-content/uploads/2012/02/166987_10150634210759250_754109249_8984775_54639276_n.jpg
#goldsborowebdevelop ment.com/wp-content/uploads/2012/02/420930_10150634211719250_754109249_8984779_2126751342_n.jpg
My cousin has the latest Intel i5 and this rig rips him apart on passmark and he's using the same hardware (I too am using 3x EVGA 570's with the same model 16gb ripjaw ram)
I have noticed that when I run the Windows test, it only loads all cores to 60%! Try it yourself download the CPU gadget for your desktop...
All CPU Meter from addgadget.com is a good one, and you will see when you run the windows test it only loads the AMD to 60%!!! What about the other 40%????
This is why its showing a low score... Trust me this CPU is one of the best multitasking cpu's on the market right now...
Also its my first build with all AMD, AMD chipset, processor and the new AMD performance memory 16gb's worth, all in all only cost me £350 for the mainboard, memory and processor and £100 for the ATI 6850 gcard and another £150 for the 256GB solid state hard drive...
If this was an I7 build it would have cost near £1000 and its not even a real 8 core processor...
What else can I add it really does rock. ;)
Asrock z77 pro-4 $110
4x4 GB ddr3 $80
Far from $1000!
You might consider shopping in some other place...
Oh yes, and it thrashes a 8150. I know, I have both.
And you need a $100 H100 or a custom loop to reach 4.6GHz with the fx, when a 2500k does it with a $30 Hyper 212.
I can't even sell my fx-8120: it's such a failure than nobody wants one! Even at 120 bucks!
It remains to be seen if this strategy will have any real-world impact, but benchmarks will certainly improve if they're multi-threaded.
I hope they pull a rabbit with Piledriver. The decision to focus on clockspeed rather than shortening the pipeline seemed like a strange plan.
not sure if i can put links in here so i'll just say its on toms hardware. Microsoft has already released the amd bulldozer hot fix which brings performance to proper levels using the (SMT) scheduling features correctly. You can gain the hotfix for windows 7 on request from microsoft.
##itproportal.com/2011/12/19/microsoft-pulls-windows-7-bulldozer-performance-patch-following-complaints/
WHAT A waste
I was trying to find the chap who quoted that his Windows 7 performance score at stock was 7.2, I do not believe in overclocking, why damage a system to get another 10% out of it. My system at stock scores 7.7 with the same for memory. So build quality can make such a difference even you should know that.
And dont take it to heart, I did not mean to offend, just could not understand why your scores were lower!
You claim you've built a system where the "processing power, GPU, and chipset have all been sourced from 1 company." Well, let's see: I was using a motherboard with an AMD 890FX chipset, and AMD FX-8150 CPU, and an AMD Radeon video card. Did you even read this review?
It just seems unusual that I am getting 10 to 15% more than your review on my ASUS board... its the M5A78L, which I will be writing a review on it later...
Obviously the Bulldozer CPUs were brand new at the time of these tests, and vendors have have more than six months since then to fine-tune their BIOS code. I imagine if I retested today with the current BIOSes the scores would be somewhat higher.
As I said somewhere else, someone said they got 7.2 on the WIn 7 check, but I am getting 7.7 at stock speeds... Might be worth a follow up, how about running it with Windows 8 so we can see what a difference we may get, as AMD say it does run better with WIN 8, however saying that I am very happy with the way my system is running, its a very nice processor which I feel will be more future proof then others. ;)
Ok how about just 1 test, which you ran before?
Have you checked to see if any updates are out for your mainboards. My bios for mine was completed end of FEB. How much performance can you get from just software and BIOS updates?
Have you ever tried it before, say with the Sandy bridge when they were slowing down...
you could expenct proper hardware support so it would be weird to ignore this and leave it out.
Imagine making a tablet version of Windows that won't run Windows software....well, you don't have to imagine any more because MS pulled it off. What a bunch of great guys. (I hope they know what they can do with their ribbon)
The new memory management/memory saving feature is intriguing.
Now we wait to see how much prices of previous generation AMD hardware drops...
The 1090T is $20 dollars cheaper than the 1100T (at Newegg), and only .1GHz slower, but it's still a Black Edition, and I'd venture that the overclocking is not that much better, if at all better, between the two. On top of this, the older releases of Phenom IIs are regularly seeing $10-$15 promo-code sales at Newegg, dropping their prices even further. $114.99 dollars for Phenom II x4 965 or (for a current example) $159.99 for 1090T is much more justifiable than spending $190 on an 1100T when ~$20-30 extra will get you a 2500K.
Future is SB-E and Ivybridge.
FX8150 is already smashed up by the 2600k.
It will be funny to see it burried by sb-e and IB!
Despite of this actual fact, I am bloody sorry, 'cos going the way it goes, Intel wil have monopoly for middle and high end CPUs within a couple of years and prices will reach the sky.
BTW, unless you are a hardcore OC'er using Ln2 and DICE, I really don't see the point of buying a FX. More watt, less power!
I love AMD and like Intel, but their BD is a FAIL. And when windows 8 will be out (I am not talking about developer version for "experts" like you, but retail, you know, for "normal" PPL), BD will be crushed by SB-E and IB. That's fact, it is already crushed by 2xxK series.
You think I like monopoly? Well,that's the end of competition, and competition is good for customers, lowering prices and bringing better products.
I think your "expertise" lacks judgement, lol.
Talk about a fanboy.
Intel has no policy about running memory > 1333MHz automatically voiding your warranty. However, running the memory voltage > 1.5V will.
However, I have to point out that if you are overclocking (as I am) it's up to YOU to take responsibility for what you're doing. Most of the time you can get away with it, but if you kill a piece of equipment while overclocking, there's no one to blame but yourself. That's the chance you take when you overclock. This is hardly Intel's fault. I know people these days don't want to accept responsibility for anything, but in this case no one else is responsible.
Also, to the reviewer, most of your graphs include an i5-"23500k OC" instead of the 2500k(?). I'm sure you were just pressed for time to get the review posted, though, haha.
The only time this ever worked for me was with the i7 920. My longest running chip between upgrades, still have it three years later, blazing at 4.2Ghz with apps that can use it's cores.
Buy what's fast today. Don't worry about tomorrow. AMD dropped the ball on BD.
"it operates 2 fpu's in a ganged mode to perform a hack job emulation of the wide instructions and cripples the cpu performance."
That post criticized FP performance when FP was one of the things it excelled in. So... yeah. That's wrong.
As a comparison, I ran the Cinebench 11.5 benchmark on my rig.
Core i7 920 3.8OC
OCZ Gold 12GB RAM at 1440Mhz
MB: Gigabyte x58 Extreme ver 2
Cinebench 11.5 score: 6.27
I ran it with many open programs taking up over 4.5GB of ram. Some of the apps I had open were Photoshop CS5, Lightroom 3, ZoneAlarm, Avast--which I shut off for the test, but left the program loaded--Open Hardware Monitor, several utility gadgets, Speedfan, Foobar 2K, Firefox 6, and several more background apps. I don't know if that would make any difference, but I just wanted to list them in case. (I couldn't close out all my apps because I had open work in them. I didn't reboot or anything. I just downloaded it and ran it.)
Well, it looks like we'll have to wait once again for the prize fight between AMD and Intel. Not a bad fight from AMD, but once again, falls short of the prize. Since 2004 and x64 AMD has gotten its ass kicked by Intel every time, except in the price range. Ouch!
With their graphics segment barely breaking even, it's going to be difficult times for AMD.
This all sounds a bit fatalist, but suffice to say the outlook is far from good.
/index.php?option=com_content&task=view&id=714&Itemid=38
Exmples of questions I am curious about: would a Single-Core AMD processor running at 2GHz be sufficient?
Would a Dual-Core Athlon processor? Would a Dual-Core Sandy Bridge? Do these lower cost parts degrade the effectiveness of high end GPUs like the GTX 580 or AMD 6970 such that those high-end GPUs practically require high-end CPUs to match?
etc, etc...
arma 2 17-37fps. the low fps are important. this unit is working like would be stuck. by the way after 4 days od having it its running on 4.1Ghz still (no tubrocore) its more stabil- in the stock relase is has ben traic. im to lazy to bring it back to shop.
Not saying that even close to 60% of my 2500k is used, just saying the old card hamstringed my video card hard.
My two bits.
FX-8150 - $259.99 (shipping estimates is between 30-45 days)
FX-8120 - $219.99 (shipping now)
FX-6100 - $189.99 (shipping now)
Then again, even newegg is out of stock of the 8150 and Amazon doesn't even have a product listing for it.
For gamers the 4170 or 4100 may be the best value, and produce the highest frame rates .
Downclock your gfx card with the driver utility (CCC for ATI° to half it's frequency and run 3dmark11. Check graphic score.
Increase the gfx card freq by 10%, run 3dmark11 and check gfx score.
Do so until you reach either a constant gfx score in 3dmark11 (cpu limited) or you max card clock (not cpu limited).
/index.php?option=com_content&task=view&id=721&Itemid=62
Think about it this way too. Bulldozer is AMDs first real change, on par with Sandy Bridge, in a while, and they are already near top performance of Intel. Now, I said near, not there. But Bulldozer is the first iteration of a whole new approach to CPU development by AMD that current OS's don't support or don't support well. (The same thing happened with SSDs and Windows, if you remember.)
How well will Bulldozer technology scale compared to Intel's? We'll just have to wait and see. AMD has been playing catch-up in the desktop speed market, and now their right on the heels of their competition.
Disappointed to say the least, especially at AMD's releasing something right out of the gate AGAIN that needs fixing and optimization to work on already existing software.
I skipped Phenom I because it was a pig and stuck with older AM2 dual core tech that did a very good job for the price, then upgraded to Phenom II when it intro'd in January '09 and was very satisfied.
Now we have a processor meant more for server applications IMO than desktop... having suffered through delays and mis-marketing only to find out it performs badly in many areas on CURRENT software is not encouraging to say the least.
AMD, I'm NOT a guinea pig...I won't buy a CPU that needs further refinement to realize it's full potential, nor will I buy a "testbed" for future technology.
Get it right like Intel did with Core2, Nehalem and Sandy Bridge, get it right like you did with the old AMD 64 stuff...but don't make statements like " will perform 50% better than i7 and Phenom II" only to have it fail like it did.
Back to Intel I go, for an i5-2500K with stellar performance for $60 US less.
Better luck next time.
I don't think AMD will ever get it right.
Intel is where it's at.
So then its not really a fair matchup then is it, lol.
I think this article did an adequate job of showing that the 8150 is barely a 2500k competitor, and in my humble opinion, far overpriced for that reason. AMD usually had the edge in performance/$. In that case, it should be priced below the 2500k, not above it. Looks like they are hoping the marketing advantage of calling it an 8-core will win over the poorly informed. At this rate, I'll be buying Ivy Bridge to replace my Phenom II.
"overclocking both the processor and memory, which only a tiny fraction of users ever do"
Be sort of stupid to buy a 2500K if your not going to OC it, even most that buy the stock 2500 use turbo charge to OC it, which is OCing. Further if it doesn't matter (OCing), why have a section of the review on just that.
"What, you want I should rag on them some more?"
Were't they advertising the Dozer as the chip that would equal or better Intel's chips, yet they are still behind Sandy Bridge, even further behind Sandy Bridge E and Ivy Bridge is around the corner. If AMD put a little more money into R&D and developed a good MC (Memory Controller) they'd have a much better product, but then too prices would go up, and as is with the Dozer the PP (price point) is already yto high for the performance delivered....which is why it is Dull
We test CPUs at their stock clock speeds with their highest officially supported memory speeds, because that's what the manufacturer guarantees they'll work at. 95+% of any CPU's sales are in pre-built systems like Dell, HP, etc. that are not only never overclocked but in most cases CAN'T be overclocked.
We overclock the CPU being tested to get an idea of the extra performance an enthusiast can unlock. Results are never guaranteed and a comparison of overclocked speeds of all CPUs might be interesting to some people but would have virtually no ral-world relevance.
And, for what it's worth, AMD never advertised the Bulldozer as something that would "equal or better Intel's chips." I spent a full day in Austin at AMD's Bulldozer tech conference and the most they said is that it would be performance-competitive with the 2500K, which it is.
No need to reply, I'm gone, I'll stick with sites that ACTUALLY perform reviews that are objective rather than w/ bias. And just a note, if you actually want to get truer results it would be best to test with sticks that are based on what you want.....i.e. if you want to test with rated 1333 sticks get a set of 1333 sticks with JEDEC rated timings, don't take high performance 2133 sticks rated for CL7 and dump them down to 1333 and raise the CL to 9, even JEDEC 8-8-8-24 is slow for these.
I think you're all wrong there...
Now, if you are talking about the Phenom II line of chips, I agree there are incredible bargains to be had. The 955 is a great value! But in terms of performance, it is not competitive. As always, its a matter of picking the components for the job at hand. If speed is what you need, Intel is the only game in town. If you want value, the OLD AMD lines are great bargains, but this new FX line is, imho, priced completely incorrectly. The FX-8150 should be cheaper than the 2500k, not more expensive.
And if it makes any difference, I am writing this from an Opteron 165, my main desktop is a Phenom II 955, and my mobile device is an E-450. But at this rate, my main desktop will be getting an Ivy Bridge update in the coming year, unless Piledriver is something that competes. That computer I use for extreme multi-tasking over 3 monitors worth of running programs, as well as Folding@home when I am away. So I have often hit with wall with my 4GHz trusty old Phenom. I'd have dropped in a FX series upgrade but my motherboard is only AM3 and does not support that chip. I'll need a new motherboard either way, so I'll most likely make it something LGA1155.
Given the growth of "the cloud", it's not surprising that Intel and AMD are putting so much effort into server processor design. But remember both companies have very strong consumer CPUs: Intel's original Sandy Bridge and forthcoming Ivy Bridge processors, and AMD's "Fusion" CPUs (even if enthusiasts sneer at them, Fusion CPUs are perfect for 90+% of the market).
All that said, I do think your idea of introducing the FX-8150 as a four-core CPU with some AMD brand of Hyper-Threading would have been a great idea for AMD.
good theory but i'm a gamer who wasn't interested in multitasking i've built a FX bulldozer 8120 Radeon HD 6950 2gb ddr5 16 gb ddr3 1666 mhz. and as it turns out since I've been able to multitask I've been using this machine to much more of its capability the thing is though i could have installed the I5 and i suspect though on a whole it doesn't perform as well as the bulldozer i wouldn't be able to see any differentiating results. I'm a gamer and by rights I should have bought an I5 however I'm also english which means i love the underdog I cannot resist an heroic second lol. It doesn't matter whether you buy a FX 8120-8150 I5 or I7 the chances are you will love the result for what you want to do just do me a favour and chuck 1000w minimum psu in and 16gb of ddr3.
Thanks for the review. When you wrote it i'd like to bet you where expecting this kind of debate and it has been fun, people are now talking about differences that they will never really notice in real world situations i like to stick to the fact that the FX8120 (which im using) is a more than adequate CPU for the price tag and has served me well so far. Cheers for this i've learnt more about CPU's in this last couple of weeks than at any other time in my life (pretty sure i could design one now lol).
Also one last thing the windows 7 update for SMT has now been released im currently running it and so far so good.
Thanks Again
I have the FX 8120 and i have to say i do use it for gaming and i find it amazing at multitasking I can run Battlefield 3 with as many programs as i want running in the background and the system has never got over 50% cpu usage.
When you say "cores", I believe you mean "threads", as there is no i7 8 cores yet (who's wrong?)...
50% on each core? How do you explain then that it run with 1 to 2 more fps with HT disabled on i7 cpu's?
Link: ##overclock.net/t/1151970/my-own-bf3-benchmark-hyperthreading-on-vs-off
- Show quoted text -
very true it is GPU intensive you are right however modern games do use a hell of alot of cpu power as well they have to network they usually need something horrible like steam to run (which eats resources) they have to handle predictive technology which lowers the lag problems these things and many more use cpu power rather than gpu then you run loads of other entertaining things on your second monitor and the draw increases. I understand what your saying but the only thing (and its a biggy) that the gpu does is run your output.
##amd.com/us/products/desktop/processors/amdfx/Pages/amdfx-model-number-comparison.aspx
##amd.com/us/products/desktop/processors/amdfx/Pages/amdfx-competitive-comparison.aspx
##amd.com/us/products/desktop/processors/amdfx/Pages/amdfx-key-architectural-features.aspx
One thing AMD have got that Intel does not have, is I have a seem less system with all AMD spec equipment, that works really well together, along with the AMD chipset and ATI 6850 graphics card, intel will never be able to do that AMD have done.
Create a cheap gaming machine for under £550 using only AMD / ATI components. ;)
My opinion would be that Intel may have to charge more because of some of this but AMD should get some credit. And most of us probably could care less about a 32 bit, (if you ask me - they should be extinct!) but where would Intel be? AMD could Rule!
Anyway- everyone thank AMD for that bad ASS I5 2500K, it should have the name AMD on it, I am sure they both deserve credit, and two good ideas made what we have now. AMD makes their own processors more than likely while Intel takes a good AMD and soups it up with their works. Maybe!.
reason I like it because it was a good price along with the motherboard and spent more on RAM and a nice full size case.for what I do I wouldnt know the difference if I had Intel except for the price of the CPU and motherboard.
Given your transition, I'd have bet you'd have been similarly pleased with a 2500.
If you're really interested, you can download the trial version of Lightwave and the Spec APC scripts and run the benchmark yourself.
VERY PLEASED with it for what I needed it for.
One more thing pad what score have you got on the WIN 7 guide and did you overclock?
I did not overclock, its fast enough as it is... ;)
Have you done that?
However, I'm puzzled by your assertion that AMD "said to use" 2133MHz memory. 2133MHz memory is hardly standard, and AMD certainly doesn't "recommend" it, since the highest DRAM frequency their 990FX chipset officially supports is 1866MHz.
Your messages would be much easier to understand with punctuation...
A year earlier I helped a friend build an intel based system with the i5 2500K processor. We used a P67 Revision 3 MB I believe. It has 8 GBs of 2.0GHz RAM but only uses a GT440 Series GPU. It can play all the games at high res with full detail just like mine. The only difference is, his system cost over 1,400 USDs to build. We searched Google Shopping, did all the comparisons and made sure to get the best price. His processor and MB were the most expensive items and were far more pricier than my MB and processor.
I believe that when it comes to personal gaming, AMD takes it with the FX-8150. It just had the price that everyone is looking for. If you are a loyal Intel fan and have the money to spend then you ARE making a good choice but for those of us who work for a living...we need those savings...period.
Point 1: AMD disagrees with you. That's why they promote the FX-8150 as an "eight core processor" rather than the "four core processor" it would be if they thought a Bulldozer module was a core as you do.
Point 2: Software disagrees with you. Single-threaded code will not magically re-arrange itself to run on the two integer execution units of a Bulldozer module; it will run on just one. Nor is there any setting or configuration you can tweak to make this happen.
No benchmark in the world will run on one Intel core and yet spread the load to run on all the resources of a Bulldozer module. And if such a benchmark did exist, it would be useless, since no other software works that way.
Regardless of the doubtless wonderful code you're cranking out, the fact is that most consumer application code remains (primarily) single-threaded. This means that single-core performance is currently the most important aspect of CPU performance.
This is why it's important to compare single-core performance of the FX-8150 against Intel processors. I don't see why this is so hard to understand.
Of course any modern OS keeps dozens of threads going, so a multi-core CPU can help system responsiveness even if individual applications don't run any faster. AMD used to beat this drum when the weakness of their single core performance relative to Intel was brought up, but I haven't seen them do it lately.
Again, just look at the benchmarks. I ran a lot of them, and the performance trend is very clear. AIDA64, CINEBENCH, SPECviewperf, SPECApc, media encoding, POVRay, Blender....all heavily multithreaded, and the FX still can't pull decisively ahead.
Performance isn't everything, of course. The FX platform has a lot more PCI-E lanes, and six native SATA 6G ports, so that's something to keep in mind.
And while the 8150 was $254 when I wrote this review, it's currently selling for $189 at Newegg...which should also tell you something: the 2500K price is actually _up_ a few dollars in that same time period.
I refuse to just stare benchmark numbers all day and conclude that one processor is better than the other without thinking about what it actually mean. And thanks for people hyping the benchmark numbers in the wrong section, great cpu like the FX-8150 doesn't get as a great of sales number.
But I do appreciate for the numbers that people should really care about, if they actually considering about usage.
There were relatively few benchmarks, though, and no single-core benchmarks. In fact we were specifically asked not to run any single core benchmarks on the demo machines they had available.
We did get the "AMD CPUs excel at multitasking" spiel. This may well be true, but the problem with this claim is that it's very hard to quantify repeatably. I can run synthetic and real-world benchmarks all day long, and get repeatable results, which I can then use to compare the performance of the items I'm reviewing. But trying to quantify something like "System responsiveness seems better with the Bulldozer when running 18 background tasks" is kinda hard. How hard? Well, when asked, AMD couldn't offer a specific test scenario that would show this in a consistent fashion.
While I'm not an AMD fanboy per se, I still use the system I built to test this processor. It's outfitted with 8GB of AMD Black Edition qualified memory and a bespoke XSPC water cooling system, all in an AMD-themed Cooler Master HAF case I reviewed a couple of years ago. It's a powerful and visually impressive system (LED lightning, don't you know). It's overclocked and much faster than most user's machines, and I have no complaints at all about its performance, nor would almost anyone else.
That still doesn't change the fact that a system based on the Intel 2500K would be just as fast if not slightly faster, and with a higher spec Intel CPU, or an overclocked on, there would be no contest.
With the price cuts on Bulldozer it arguably has more bang for the buck than Intel, and things like the extra PCI-E lanes and SATA 6G ports a Bulldozer system offers are other things to consider. But there's no point in pretending that Intel doesn't decisively own the performance crown.
For an all round gaming package you cannot beat an AMD machine. And its a lot cheaper to!
Gee, if only someone had run 27 or so synthetic and real-world benchmarks comparing the 8-core 8150 against a four-core Intel chip. Then we could look at actual numbers and make an informed decision.