 |
 |
 |
||
|
|
||||
 |
|
|
| Introduction and Specifications | ||||||||||||||||||||||||||||||||||||||||||
Save for its unique markings, the new AMD Phenom X4 9850 looks exactly like any other socket AM2+ processor outfitted with AMD's standard heat spreader. The chip uses the same packaging and socket as current Phenom processors; it is only the silicon underneath that has changed. |
| Vital Signs and Overclocking | ||||
As we've mentioned, AMD’s new Phenom X4 9850 processor look just like current Phenom processors. It utilizes the same packaging and pin configuration. But underneath that unassuming heat-spreader is a new silicon revision that fixes the TLB bug present on initial Phenom processors.
The latest version of CPU-Z correctly identifies and outlined the Phenom X4 9850's inner workings. As you can see, the processor is based on the Agena core with socket AM2+ packaging. It is clocked at 2.5GHz (12.5 x 200MHz), although the clock speed is reported as 1255MHz due to AMD's Cool'n'Quiet feature being enabled. Our particular chip is stepping ‘DR-B3’ and supports all of the instructions sets listed in the aptly names ‘instructions’ section. The processor’s 2MB L2 cache configuration (512KB per core) is 16-way set associative, but the processors 2MB of shared L3-cache is 32-way set associative. With the exception of the stepping and clock speed, nothing else has changed with this revision of Phenom.
We know many of you are wondering just how much clock speed headroom the Phenom X4 9850 has left under the hood, so we spent some time overclocking our chip using a Gigabyte 790FX-chipset based motherboard. We simply upped the processor voltage and increased the multiplier until our test machine was no longer stable. A basic Thermaltake air cooler, similar to AMD's stock PIB heatsink was used.
With a minor bump in voltage up to 1.4v, we were able to take our Phenom X4 9850 processor up from its default clock speed of 2.5GHz to a little over 2.8GHz, an increase of about 300MHz or 12%. We suspect that with more elaborate cooling and perhaps a little more voltage, even higher frequencies are possible. And remember, because this is a Black Edition processors, anyone who buys one of these chips at retail will also be able to alter its multiplier exactly as we did here. AMD's excellent Overdrive software can also be used to overclock the chip, for those less inclined to fiddle around in the system BIOS. |
| Our Test Systems and SANDRA | |||||||||||||||||
How We Configured Our Test Systems: When configuring our test systems for this article, we first entered their respective system BIOSes and set each board to its "Optimized" or "High performance Defaults". We then saved the settings, re-entered the BIOS and set memory timings for either DDR2-1066 (AMD) with 5,5,5,15 timings or DDR3-1333 with 7,7,7,20 timings (Intel). The hard drives were then formatted, and Windows Vista Ultimate was installed. When the Windows installation was complete, we updated the OS, and installed the drivers necessary for our components. Auto-Updating and Windows Defender were then disabled and we installed all of our benchmarking software, defragged the hard drives, and ran all of the tests.
|
| PCMark Vantage | ||||
We ran a handful processors and platforms, including the new Phenom X4 9850, through Futuremark’s latest system performance metric built especially for Windows Vista, PCMark Vantage. PCMark Vantage runs through a host of different usage scenarios to simulate different types of workloads including High Definition TV and movie playback and manipulation, gaming, image editing and manipulation, music compression, communications, and productivity. Most of the tests are multi-threaded as well, so the tests can exploit the additional resources offered by a quad-core CPU.
From this point forward in this article, we'll be comparing the performance of the new AMD Phenom X4 9850 processor to a couple of Intel quad-core processors and a trio of other Phenom configurations. To get as close to an apples to apples comparison between the B3 and B2 revision of Phenom, we overclocked a Phenom 9600 Black Edition to 2.5GHz and ran it though our benchmark suite with the TLB patch both enabled and disabled. |
| LAME MT and Kribibench v1.1 | ||||||||
In our custom LAME MT MP3 encoding test, we convert a large WAV file to the MP3 format, which is a popular scenario that many end users work with on a day-to-day basis to provide portability and storage of their digital audio content. LAME is an open-source mid to high bit-rate and VBR (variable bit rate) MP3 audio encoder that is used widely around the world in a multitude of third party applications.
In this test, we created our own 223MB WAV file (a hallucinogenically-induced Grateful Dead jam) and converted it to the MP3 format using the multi-thread capable LAME MT application in single and multi-thread modes. Processing times are recorded below, listed in seconds. Once again, shorter times equate to better performance.
For this next batch of tests, we ran Kribibench v1.1, a 3D rendering benchmark produced by the folks at Adept Development. Kribibench is an SSE aware software renderer where a 3D model is rendered and animated by the host CPU and the average frame rate is reported. We used two of the included models with this benchmark: a "Sponge Explode" model consisting of over 19.2 million polygons and the test suite's "Ultra" model that is comprised of over 16 billion polys.
Once again, it is clear that the Phenom X4 9850 is AMD's fastest desktop processor to date. It outpaces all of the other AMD-built chips to some degree, and just edges out Intel's Core 2 Quad Q6600. |
| Cinebench R10 and 3DMark06 | ||||||||
Cinebench R10 is an OpenGL 3D rendering performance test based on Cinema 4D. Cinema 4D from Maxon is a 3D rendering and animation tool suite used by 3D animation houses and producers like Sony Animation and many others. It's very demanding of system processor resources and is an excellent gauge of pure computational throughput.
This is a multi-threaded, multi-processor aware benchmark that renders a single 3D scene and tracks the length of the entire process. The time it took each test system to render the entire scene is represented in the graph below, listed in seconds.
3DMark06's built-in CPU test is a multi-threaded DirectX gaming metric that's useful for comparing relative performance between similarly equipped systems. This test consists of two different 3D scenes that are processed with a software renderer that is dependent on the host CPU's performance. Calculations that are normally reserved for your 3D accelerator are instead sent to the CPU for processing and rendering. The frame-rate generated in each test is used to determine the final score.
3DMark06's CPU performance module tells essentially the same story as Cinebench. The Phenom X4 9850 is slightly faster than the similarly clocked B2 revision chip, and just a notch behind the Core 2 Duo Q6600. |
| Gaming: Crysis and F.E.A.R. | ||||
For our next set of tests, we moved on to some in-game benchmarking with Crysis and F.E.A.R. When testing processors with Crysis or F.E.A.R., we drop the resolution to 800x600, and reduce all of the in-game graphical options to their minimum values to isolate CPU and memory performance as much as possible. However, the in-game effects, which control the level of detail for the games' physics engines and particle systems, are left at their maximum values, since these actually do place some load on the CPU rather than GPU.
In both F.E.A.R. and Crysis, the new AMD Phenom X4 9850 processor was somewhat faster than the similarly clocked B2 revision chip. Also note the huge performance penalty associated with running a B2 revision chip with the TLB patch enabled. With the patch enabled a Phenom 9600, clocked 200MHz below the 2.5GHz chip is actually faster to a large degree. In comparison to the Intel processors, the Phenom X4 9850 smokes the Q6600 in F.E.A.R., but falls behind in Crysis. |
| Total System Power Consumption | ||||
Before we bring this article to a close, we'd like to cover a few final data points. Throughout all of our benchmarking and testing, we monitored how much power our test systems were consuming using a power meter. Our goal was to give you all an idea as to how much power each configuration used while idling and under a heavy workload. Please keep in mind that we were testing total system power consumption at the outlet here, not just the power being drawn by the motherboards alone.
|
| Our Summary and Conclusion | ||||
Performance Summary: The Phenom X4 9850 is AMD's most powerful desktop processor to date. Due to the processor's higher clock speed and revised silicon that eliminates the TLB errata, the Phenom X4 9850 is measurably faster than all of AMD's previously released Phenom processors. In comparison to Intel's closest competitor, the Phenom X4 9850 competes well, trading benchmark victories with the 2.4GHz Core 2 Quad Q6600. Keep in mind though, Intel is updating their line-up to include Penryn-based Core 2 products across the board and higher performing, 45nm Yorkfield-based quad-core processors should be hitting store shelves very soon. In fact, the Q9300 has already showed up at some on-line retailers for about $290.
Although the Phenom X4 9850 doesn't propel AMD to the head of the pack in terms of performance, it is a significant step forward for the company. AMD can now put the TLB issue behind them and focus on ramping clock speeds and their impending transition to a 45nm manufacturing process. Until all that happens, the Phenom X4 9850 allows AMD to compete better with Intel's offerings and eliminates a roadblock that was preventing some AMD faithfuls from migrating to a Phenom CPU. And when you consider the total platform, AMD is in a pretty good position. DDR2 RAM is dirt cheap at the moment, and AMD 770 and 790FX motherboards are priced anywhere between $100 and $220, not to mention the affordable 780G. In addition, the list of older socket AM2 motherboards that support Phenom is growing all the time. Drop in CPU upgrades sure are nice when a user doesn't have the funds to overhaul their entire system.
|
It doesn't take a genius to figure out that AMD's initial Phenom processor launch didn't go exactly as planned. The first batch of retail-ready Phenom processors were clocked lower than expectations and in the days leading up to the launch, the now infamous TLB bug reared its ugly head. If you're not aware, all initial Phenom and Opteron quad-core processors based on the Barcelona architecture were plagued by an errata that could cause a system crash under certain circumstances. The errata involves the processor's translation lookaside buffer or "TLB" and L3 cache. Keep in mind, every processor has some sort of errata, but it usually isn't cause for concern, nor does it manifest itself in any real-world application performance. In many cases processor errata can be worked-around via BIOS or software patches and end users are none the wiser. With the Phenom's TLB errata though, working around the problem involved a significant performance penalty.
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
;)
