Athlon II X4 635 and Phenom II X2 555 CPU Review

[nextpage title=”Introduction”]

Today we are going to see the performance of two new CPUs from AMD, Athlon II X4 635 (2.9 GHz) and Phenom II X2 555 (3.2 GHz), and compare it to other models from AMD and competitors from Intel.

On AMD product line, “X2” means that the CPU has two processing cores (“dual-core”), “X3” means the CPU has three processing cores (“triple-core”) and “X4” means that the CPU has four processing cores (“quad-core”).

The main difference between Athlon II and Phenom II is the presence of an L3 memory cache on Phenom II, making this CPU to provide a higher performance when compared to an Athlon II of same clock rate and same amount of processing cores. Both Athlon II and Phenom II are socket AM3 CPUs, meaning that they can be installed on socket AM2+ or socket AM3 motherboards, depending on whether you want DDR2 or DDR3 memories on your computer, respectively.

The new Phenom II X2 555 (3.2 GHz) comes with an attractive USD 99 price tag for a CPU on this clock range. In fact it comes with the exact same price tag as Athlon II X4 620 (2.6 GHz), a CPU that we’ve already reviewed and proved to provide a good cost/benefit ratio. So we are going to compare these two CPUs to answer the big question: since the two CPUs cost the same, which one is the best? The CPU with the higher clock rate and an L3 memory cache but only two cores or the CPU without an L3 cache and lower clock rate, but with four processing cores?

As a competitor to these two CPUs, we elected Pentium E6300, which is offered at USD 80. Since motherboards targeted to Intel systems are more expensive and since the difference in price between the motherboards we used on our review is exactly USD 20, we ended up with systems with the exact same price tag.

We see all the time other websites publishing flawed reviews, for instance comparing CPUs that are in completely different price range (e.g., comparing these CPUs from AMD with Core i3 or Core i5 models). Even when you compare CPUs on the same price range the comparison is flawed, because you have to put the cost of the motherboard into the equation, since you have to compare the price of the whole system, not only the price of the CPU alone.

The second AMD CPU we are reviewing is Athlon II X4 635 (2.9 GHz), which is being offered at a higher price point (USD 126). Intel doesn’t have an “exact” competitor to this CPU, so we decided to pick as competitor Core 2 Duo E7500 (2.93 GHz), which costs around USD 116 and was the closest match. This comparison should be really interesting because both CPUs run at almost the same clock rate, with the AMD unit having as advantage having four processing cores, while the Intel counterpart has as advantage having SSE4 instructions.

AMD is targeting these CPUs to be paired with motherboards with integrated video based on AMD785G. Because of that we had to pick a motherboard with integrated graphics on the Intel side as well (Intel G45). Of course several users will install the CPUs we selected on system without integrated graphics. So in order to cover this situation, we conducted two tests with each CPU: with on-board graphics and with a GeForce 9600 GT installed.

Before going to our results, let’s compare the main specs from the CPUs included in this review.

[nextpage title=”Main Specs”]

Architecture-wise the main difference between the CPUs from AMD and their Intel counterparts is the presence of an integrated memory controller that supports both DDR2 and DDR3 memories on AMD processors. Intel Core i3, Core i5 and Core i7 like AMD CPUs have an integrated memory controller, however these CPUs are way more expensive than the reviewed CPUs from AMD and thus were not included in this review. Therefore memory support on Core 2 Duo and Pentium CPUs depends on the chipset. Since Intel G45 chipset only supports DDR2 up to 800 MHz, we used the same DDR2-800 modules with all CPUs, this way the memory modules wouldn’t bias the results.

From the CPUs included in our review, only Core 2 Duo E7500 supports SSE4.1 instructions (but not SSE4.2). AMD CPUs have AMD’s own implementation of SSE4 called “SSE4a,” which has only four instructions and is not the same thing as SSE4 (which has a total of 54 instructions, 47 on SSE4.1 and seven on SSE4.2).

In the table below there is a column called “external bus.” For the AMD CPUs this means the speed of the HyperTransport bus, which on the reviewed CPUS works at 2,000 MHz, which translates in a maximum theoretical transfer rate of 8,000 MB/s (2,000 MHz x 16 bits x 2 data per clock cycle / 8). On Intel CPUs the external bus is called front side bus (FSB), which on the selected CPUs works at 266.6 MHz transferring four 64-bit data per clock cycle. Because of that, this bus is also referred as “1,066 MHz” (4x 266.6 MHz). This is equivalent of a maximum theoretical transfer rate of 8,533 MB/s (266.6 MHz x 64 bits x 4 data per clock cycle / 8).

CPU	Cores	Internal Clock	External Bus	Core	Tech.	TDP	Socket	SSE4	Price
Athlon II X4 620	4	2.6 GHz	8 GB/s	Propus	45 nm	95 W	AM3	SSE4a	USD 99.00
Athlon II X4 635	4	2.9 GHz	8 GB/s	Propus	45 nm	95 W	AM3	SSE4a	USD 126.00
Phenom II X2 555	2	3.2 GHz	8 GB/s	Callisto	45 nm	80 W	AM3	SSE4a	USD 99.00
Pentium E6300	2	2.8 GHz	8.5 GB/s	Wolfdale	45 nm	65 W	775	No	USD 80
Core 2 Duo E7500	2	2.93 GHz	8.5 GB/s	Wolfdale	45 nm	65 W	775	SSE4.1	USD 116.00

TDP stands for Thermal Design Power which advises the user of the maximum amount of heat the CPU can dissipate. The CPU cooler must be capable of dissipating at least this amount of heat.

The prices listed were researched at Newegg.com on the day we published this review, except the price for Phenom II X2 555, since we couldn’t find this CPU being sold yet on the day we published this review (in this case the price is for distributors in the USA buying 1,000-unit lots; the actual street price will probably be a little bit higher).

CPU	L1 Cache	L2 Cache	L3 Cache
Athlon II X4 620	64 KB + 64 KB per core	512 KB per core	No
Athlon II X4 635	64 KB + 64 KB per core	512 KB per core	No
Phenom II X2 555	64 KB + 64 KB per core	512 KB per core	6 MB
Pentium E6300	32 KB + 32 KB per core	2 MB total	No
Core 2 Duo E7500	32 KB + 32 KB per core	3 MB total	No

Now that you know the contenders, let’s see how they performed.

[nextpage title=”How We Tested”]During our benchmarking sessions, we used the configuration listed below. Between our benchmarking sessions the only variable was the CPU being tested and the motherboard, which had to be replaced to match the different CPU sockets.

Hardware Configuration

• Motherboard (Socket LGA775): Intel DG45ID (IDG4510H.86A.0113.2009.0804.1430 BIOS)
• M
  otherboard (Socket AM3): Biostar TA785GE 128M (10/07/09 BIOS)
• CPU Cooler (Socket LGA775): Intel stock
• CPU Cooler (Socket AM3): AMD stock
• Memory: Two 1 GB Crucial CT12864AA800 modules (DDR2-800/PC2-6400, CL6, 1.8 V), configured at 800 MHz
• Hard Disk Drive: Western Digital Caviar SE16 500 GB (WD5000AAKS, SATA-300, 7,200 rpm, 16 MB buffer)
• Video Card: Zotac GeForce 9600 GT Synergy Edition
• Video Monitor: Samsung Syncmaster 932BW
• Power Supply: OCZ StealthXStream 400 W
• Optical Drive: Lite-On LH-20A1L

Operating System Configuration

• Windows 7 Ultimate 64-bit
• NTFS
• Video resolution: 1440×900 60 Hz

Driver Versions

• NVIDIA video driver version: 195.62
• Intel Inf chipset driver version (Intel): 9.1.1.1025
• Intel video driver version (Intel): 15.15.8.64.2020
• IDT audio driver version (Intel): 6224.3
• Video drivers (AMD): 3.00.0732
• Chipset drivers (AMD): 8.61
• Network drivers (AMD): 7.003
• Realtek audio driver version (AMD): R2.36

Software Used

• PCMark Vantage Professional 1.1.0
• VirtualDub 1.9.5 + MPEG-2 Plugin 3.1 + DivX 6.8.5
• Adobe Photoshop CS4 Extended + GamingHeaven Photoshop Benchmark V3
• Adobe After Effects CS4
• WinRAR 3.92
• Cinebench 10
• 3DMark06 Professional 1.1.0
• 3DMark Vantage Professional 1.0.1
• Half-Life 2: Episode Two – Patch June 23th 2009 + HardwareOC Half-Life 2 Episode Two Benchmark Tool 1.2.0.0
• Fallout 3 – Patch 1.7

Error Margin

We adopted a 3% error margin; thus, differences below 3% cannot be considered relevant. In other words, products with a performance difference below 3% should be considered as having similar performance.

[nextpage title=”PCMark Vantage”]

PCMark Vantage simulates the use of real-world applications and gives scores for the following categories:

• PCMark
• Memories
• TV and Movies
• Gaming
• Music
• Communications
• Productivity
• HDD

For a detailed description of each one of these tests, please download and read the PCMark Vantage Reviewer’s Guide.

You can see the results for each category below. We are not going to compare the results for the Memories and HDD suites.

Comparing the results achieved using the on-board video, Phenom II X2 555 achieved the same performance level as Athlon II X4 620, being 25% faster than Pentium E6300. Athlon II X4 635 was 11% faster than its main competitor, Core 2 Duo E7500.

When we installed a GeForce 9600 GT the performance difference increased, but Phenom II X2 555 was still at the same performance level as Athlon II X4 620, being 10% faster than Pentium E6300. Athlon II X4 635 was 12% faster than Core 2 Duo E7500.

On the TV and Movies benchmark with all systems using their on-board video, Phenom II X2 555 and Pentium E6300 achieved exactly the same performance, with Athlon II X4 620 being 23% faster. Here Athlon II X4 635 was 16% faster than Core 2 Duo E7500.

With a GeForce 9600 GT installed, Athlon II X4 635 was 35% faster than Core 2 Duo E7500, and Phenom II X2 555 was 9% faster than Pentium E6300, but Athlon II X4 620 was 26% faster than Phenom II X2 555.

On the Gaming set with the on-board video enabled, Phenom II X2 555 and Athlon II X4 620 achieved the same performance level, with Phenom II X2 555 being 20% faster than Pentium E6300. Athlon II X4 635 was 23% faster than Core 2 Duo E7500.

When we installed our GeForce 9600 GT, Athlon II X4 620 was 12% faster than Phenom II X2 555, which in turn was 10% faster than Pentium E6300. Athlon II X4 635 was 16% faster than Core 2 Duo E7500.

On the Music benchmark with the on-board video enabled Phenom II X2 555 was 12% faster than Athlon II X4 620 and 7% faster than Pentium E6300. Here Core 2 Duo E7500 was 6% faster than Athlon II X4 635.

When we installed a GeForce 9600 GT, Phenom II X2 555 achieved the same performance level as Pentium E6300, being 6% faster than Athlon II X4 620. Core 2 Duo E7500 was only 3% faster than Athlon II X4 635 under this scenario.

On the Communications tests with on-board video enabled Phenom II X2 555 was 4% faster than Athlon II X4 620 and 11% faster than Pentium E6300.  Athlon II X4 635 was 15% faster than Core 2 Duo E7500.

When we installed a GeForce 9600 GT Athlon II X4 635 was 12% faster than Core 2 Duo and Phenom II X2 555 achieved the same performance level as Athlon II X4 620, being 13% faster than Pentium E6300.

And finally on the Productivity benchmark Athlon II X4 620 was 10% faster than Phenom II X2 555, which in turn was 10% faster than Pentium E6300. Athlon II X4 635 was 20% faster than Core 2 Duo E7500 on this test.

When we installed a GeForce 9600 GT Athlon II X4 620 was 9% faster than Phenom II X2 555, which in turn was 27% faster than Pentium E6300. Athlon II X4 635 was 16% faster than Core 2 Duo E7500.

[nextpage title=”VirtualDub + DivX”]

With VirtualDub we converted a full-length DVD movie to DivX format and saw how long it took for this conversion to be completed. The DivX codec is capable of recognizing and using not only more than one CPU (i.e., more than one core), but also the SSE4 instruction set (feature not available on the reviewed CPUs).

The movie we chose to convert was Star Trek – The Motion Picture: Director’s Cut. We copied the movie to our hard disk drive with no compression, so the final original file on our HDD was 6.79 GB. After compressing it with DivX, the final file was only 767.40 MB, which is quite remarkable.

The results below are given in seconds, so the lower the better.

On DivX encoding using the on-board video Phenom II X2 555 was only 4% faster than Athlon II X4 620 and 20% faster than Pentium E6300. Athlon II X4 635 was 18% faster than Core 2 Duo E7500, which is impressive, since Core 2 Duo supports SSE4 instructions.

When we installed a GeForce 9600 GT performance increased less than 3% compared to the results taken with the on-board video, so we have to consider that there was no significant performance increase when we installed a “real” video.

[nextpage title=”Photoshop CS4″]

The best way to measure performance is by using real programs. The problem, though, is creating a methodology using real software that provides accurate results. For Photoshop CS4, there is a methodology created by the folks at GamingHeaven that is very accurate. Their script applies a series of 15 filters to a sample image, and we wrote down the time taken for each filter to run. At the end, we have the results for each individual filter and we simply added them up to have the total time taken to run the 15 filters from the GamingHeaven batch. The results below are given in seconds, so the lower the number the better.

On Photoshop CS4 Phenom II X2 555 was 22% faster than Athlon II X4 620 and 9% faster than Pentium E6300. It also achieved the same performance level as Core 2 Duo E7500, which was 13% faster than Athlon II X4 635. These results are with the on-board video enabled.

When we installed a GeForce 9600 GT performance increased less than 3% compared to the results taken with the on-board video, so we have to consider that there was no significant performance increase when we installed a “real” video card.

[nextpage title=”After Effects CS4″]

After Effects is a very well-known program for video post-production that is used to add animation and visual effects in videos. To evaluate the performance of each CPU running this program, we ran a workload consisting of 25 compositions that applied several filters and effects to a variety of input file types such as PSD (Photoshop), AI (Illustrator), EPS, and TIF. After each filter was applied, the composition was rendered to an uncompressed AVI file with the same resolution as the input files. The results below are the time each CPU took to finish the whole batch, given in seconds, so the lower the number the better.

On After Effects CS4 with on-board video enabled, Athlon II X4 620 was 29% faster than Phenom II X2 555, and Pentium E6300 was 21% faster than Phenom II X2 555. Under this scenario Core 2 Duo E7500 was 26% faster than Athlon II X4 635.

Installing a GeForce 9600 GT improved the performance for some CPUs. Phenom II X2 555 was now 33% faster, being 6% faster than Athlon II X4 620 and 8% faster than Pentium E6300. The installation of this video card improved Athlon II X4 635 performance by 35%, making it to be 12% faster than Core 2 Duo E7500.

[nextpage title=”WinRAR”]

We measured the time each CPU took to compress five high-resolution 48-bit uncompressed TIF images, each one with around 70 MB, to RAR format with the popular WinRAR application. The results below are given in seconds, so the lower the number the better.

On file compression using WinRAR Phenom II X2 555 was 8% faster than Pentium E6300 and 6% faster than Athlon II X4 620. Athlon II X4 635 and Core 2 Duo E7500 achieved the same performance level.

With a GeForce 9600 GT installed, Phenom II X2 555 was 10% faster than Pentium E6300 and 5% faster than Athlon II X4 620, with Athlon II X4 635 and Core 2 Duo E7500 achieving the same performance level once again.

[nextpage title=”Cinebench 10″]Cinebench 10 is based on the 3D software, Cinema 4d. It is very useful to measure the performance gain given by having more than one CPU installed on the system when rendering heavy 3D images. Rendering is one area in which having more than one CPU helps considerably, because usually, rendering software recognizes several CPUs. (Cinebench, for instance, can use up to 16 CPUs.)

Since we were interested in measuring the rendering performance, we ran the test called “Rendering x CPUs,” which renders a “heavy” sample image using all available CPUs (or cores – either real or virtual, as on CPUs with Hyper-Threading technology, each core is recognized as two cores by the operating system) to speed up the process.

On Cinebench Athlon II X4 620 was 35% faster than Phenom II X2 555, while Phenom II X2 555 was 13% faster than Pentium E6300. Athlon II X4 635 was 54% faster than Core 2 Duo E7500. Impressive.

Installing a “real” video card didn’t impact the results.

[nextpage title=”3DMark06 Professional”]

3DMark06 measures Shader 3.0 (i.e., DirectX 9.0c) performance. We ran this software under its default configuration. The results you can see below.

Using the on-board video, Phenom II X2 555 was 4% faster than Athlon II X4 620 and 46% faster than Pentium E6300, while Athlon II X4 635 was 39% faster than Core 2 Duo E7500. Of course these results reflect the superiority of AMD785G over Intel G45.

When we installed a GeForce 9600 GT to measure the performance of the CPU alone, Athlon II X4 620 was 4% faster than Phenom II X2 555, which achieved the same performance level of Pentium E6300. Under this scenario Athlon II X4 635 was 7% faster than Core 2 Duo E7500.

[nextpage title=”3DMark Vantage Professional”]

3DMark Vantage measures Shader 4.0 (i.e., DirectX 10) gaming performance. First we tried to run this program under the “Entry” profile with the motherboard on-board video enabled. Only AMD785G could run this program (achieving an Entry score of E2311); Intel G45 could not run 3DMark Vantage even on the “Entry” profile selected. This is old news in the industry: while officially Intel G45 is labeled as being a DirectX 10 part, in fact it can’t run DirectX 10 games or simulations.

The numbers you see below are the results running the “Performance” profile with the GeForce 9600 GT installed.

Under the “Performance” profile from 3DMark Vantage with a GeForce 9600 GT installed Phenom II X2 555 achieved the same performance as Athlon II X4 620 and Pentium E6300. Athlon II X4 635 also has achieved the same performance level as Core 2 Duo E7500. This shows that here the video card is more important than the CPU for DirectX 10 gaming.

Comparing exclusively the CPU benchmark results from 3DMark Vantage, Athlon II X4 620 was 13% faster than Phenom II X2 555, which achieved the same performance level of Pentium E6300. Athlon II X4 635 was 16% faster than Core 2 Duo E7500 under this benchmark. But as you can see, a higher CPU performance may not translate into a higher gaming performance nowadays.

[nextpage title=”Half-Life 2: Episode Two”]

Half-Life 2 is a popular franchise and we benchmarked the video cards using Episode Two with the aid of HOC Half-Life 2 Episode Two benchmarking utility using the “HOC Demo 1” provided by this program. We ran the game under 1024×768 with no anti-aliasing and bilinear filtering, i.e., using the lowest image quality possible. The results, given in frames per second, you see below.

Here Phenom II X2 555 + AMD785G was 4% faster than Athlon II X4 620 and 55% faster than Pentium E6300 + Intel G45. Athlon II X4 635 + AMD785G was 49% faster than Core 2 Duo E7500 + Intel G45.

Of course these numbers represent the performance of the motherboard, not the CPU.

When we installed a GeForce 9600 GT Phenom II X2 555 was the fastest CPU (probably due to its higher clock rate), being 49% faster than Pentium E6300, 14% faster than Athlon II X4 620, 21% faster than Athlon II X4 635 and 21% faster than Core 2 Duo E7500. Core 2 Duo E7500 and Athlon II X4 635 achieved the same performance on this game.

[nextpage title=”Fallout 3″]

Fallout 3 is based on the same engine used by The Elder Scrolls IV: Oblivion, and it is a DirectX 9.0c (Shader 3.0) game. To measure performance, we used the FRAPS utility running an outdoor scene at God mode, running through enemy fire, triggering post processing effects, and ending with a big explosion in front of Dupont Circle. First we tried to run this program at 1440×900 with all image quality settings at “low” with the motherboard on-board video enabled. Only AMD785G could run this program (achieving 20.35 frames per second); Intel G45 could not run Fallout 3. Then we installed a GeForce 9600 GT and ran this program with image quality set to “high.” The results for this scenario are presented below.

On Fallout 3 with a GeForce 9600 GT installed Phenom II X2 555 was 11% faster than Athlon II X4 620 and 9% faster than Pentium E6300. And Core 2 Duo E7500 and Athlon II X4 635 achieved the same performance level.

[nextpage title=”Conclusions”]

If right now Intel is the sole leader on the high-end market, at least on the mainstream arena AMD is nicely catching up. Depending on the kind of programs you run, you will better off picking an AMD solution.

For example, in gaming, Phenom II 555 X2 was the best CPU, easily beating CPUs that are 25% more expensive. This happens because of the higher clock rate of this CPU, combined with the presence of an L3 memory cache – keep in mind that most games don’t use more than two CPU cores. But of course where the amount of CPU cores matter (3D rendering and video editing), it is better to pick a quad-core CPU.

So if you have around USD 100 to spend on a CPU, our recommendation is the new Phenom II X2 555, unless you really need a quad-core machine. In this case Athlon II X4 620 is the perfect pick, since it costs the same and will provide you the extra performance needed.

But if you have USD 25 more to spend on a CPU, the new Athlon II X4 635 is definitely our recommendation, especially if you work with video editing and/or 3D rendering. It only lost to Core 2 Duo E7500 on Photoshop CS4, on all other programs it was either faster or achieved the same performance level. For the regular user that wants a fast machine and won’t run professional video editing or 3D rendering programs, we think Phenom II X2 555 provides a better cost/benefit ratio though.

By the way. Pay attention that this time we tested all CPUs using Windows 7, so some results may be different from our previous reviews.