Celeron, Pentium Dual Core and Athlon X2: Which One is the Best USD 70 CPU?

[nextpage title=”Introduction”]

Today we put together a review comparing the performance of entry-level CPUs on the USD 70 price range: Celeron E1400 (dual-core 2 GHz), Pentium Dual Core E2160 (1.8 GHz), Pentium Dual Core E2180 (2 GHz) and Athlon X2 4600+ (2.4 GHz). With this review we wanted to answer a simply question: which one is the best USD 70 CPU today? We also included on this round-up a Core 2 Duo E4400 (2 GHz). We know that this CPU isn’t on this price range (and not even sold in the US anymore), but we were curious to compare a Celeron, a Pentium Dual Core and a Core 2 Duo all running under the same clock rate (2 GHz in our case) to see what is the performance difference between these Intel CPUs. Which CPU is the winner? Read on!

Let’s first talk about the differences between the reviewed CPUs. In order to facilitate our explanations, we compiled the table below.

CPU	Cores	Internal Clock	External Clock	L1 Cache	L2 Cache	Technology	TDP
Athlon X2 4600+	2	2.4 GHz	*	128 KB x2	512 KB x 2	90 nm	65 W
Celeron E1400	2	2 GHz	800 MHz	64 KB x2	512 KB x 1	65 nm	65 W
Pentium Dual Core E2160	2	1.8 GHz	800 MHz	64 KB x2	1 MB x 1	65 nm	65 W
Pentium Dual Core E2180	2	2 GHz	800 MHz	64 KB x2	1 MB x 1	65 nm	65 W
Core 2 Duo E4400	2	2 GHz	800 MHz	64 KB x2	2 MB x 1	65 nm	65 W

* 200 MHz base clock and 1,000 MHz (4,000 MB/s) HyperTransport link.

All Intel CPUs included in our review are based on the same basic chip, codenamed Conroe. The difference between them is the amount of L2 memory cache that was added.  In other words, the difference between Celeron E1000, Pentium Dual Core and Core 2 Duo is the amount of L2 memory cache. So a secondary goal in our review is to see the performance impact when increasing the L2 memory cache, as the models we included were running under the same clock rate (2 GHz).

It is important to remember that Celeron is a name that has been used on processors with several different internal architectures. Celeron from E1000 series are dual-core CPUs based on the same architecture as Core 2 Duo, as explained. Celeron processors from other series are completely different animals. For example Celeron CPUs from 400 series are also based on the same architecture but having only one processing core instead of two. And Celeron D is a single-core CPU based on Pentium 4’s architecture (“D” standards for “Desktop” and not “Dual-core”).

As we mentioned, Core 2 Duo is a CPU that does not compete with the other CPUs included in this review, but since we had a 2 GHz (E4400) model available, we were curious to compare three CPUs with the same architecture and running under the same clock rate, with the only difference between them being the size of the L2 memory cache. E4400 isn’t available in the North American market anymore and today the cheapest Core 2 Duo available is E4500, which runs at 2.2 GHz and is sold around USD 125. Keep also in mind that these Core 2 Duo processors with lower clock rates have an external clock rate of 800 MHz while more expensive models have an external clock rate of at least 1,066 MHz.

The Intel CPUs included in our review has only one L2 memory cache, which is shared by the two cores from the CPU. Athlon X2 uses two L2 memory caches, one for each core.

AMD CPUs have an embedded memory controller. On Intel CPUs the memory controller is located in the chipset (in the north bridge chip, to be more specific). Athlon X2 has a problem where depending on the model the memory won’t run under their maximum clock rate (read here a full description of this issue; affected models are BE-2300, BE-2350, 3600+, 4000+, 4200+, 4400+, 4800+ with 2.5 GHz clock, 5000+ and 6000+). Luckily Athlon X2 4600+ doesn’t suffer from this problem and with this CPU DDR2-800 memories can really run at 800 MHz. Speaking of which, during our reviews we set our memories to always run at 800 MHz with the same timings (5-5-5-18) regardless of the CPU being tested. This is very important to check on a round-up review like this, otherwise some CPU can achieve a higher score than another just because memory was configured differently (different clock rate or different timings).

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.

AMD CPUs are based on socket AM2+ while Intel CPUs are based on socket LGA775.

And as a final note, AMD has recently changed the name “Athlon 64 X2” to simply “Athlon X2,” and throughout this review we will be using this new name.

[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, as AMD and Intel CPU’s require different motherboard types.

Hardware Configuration

• Motherboard (Intel): ASUS P5K-E/WiFi-AP (1013 BIOS)
• Motherboard (AMD): ASUS M3A32-MVP DeLuxe/WiFi-AP (0603 BIOS)
• Memory: 2 GB Corsair Dominator TWIN2X2048-8500C5D (DDR2-1066/PC2-8500 with 5-5-5-15 timings), configured at 800 MHz with 5-5-5-18 timings.
• Hard Disk Drive: Seagate Barracuda 7200.10 160 GB (ST3160815AS, SATA-300, 7,200 rpm, 8 MB buffer)
• Video Card: GeForce 8500 GT
• Video resolution: 1440×900 75 Hz
• Video Monitor: Samsung Syncmaster 932BW
• Power Supply: OCZ ProXStream 1000 W
• CPU Cooler: Thermaltake TMG i1
• Optical Drive: LG GSA-H54N

Software Configuration

• Windows Vista Ultimate 32-bit
• Service Pack 1

Driver Versions

• NVIDIA video driver version: 175.19
• Intel Inf chipset driver version: 8.3.1.1009
• AMD chipset driver version: 3.0.642.0

Software Used

• PCMark Vantage Professional 1.0.0
• VirtualDub-MPEG2 1.6.19 Build 24586 + DivX 6.8.3
• RazorLame 1.1.5a + LAME 3.9.7
• GamingHeaven Photoshop Benchmark V2
• Cinebench 10
• 3DMark06 Professional 1.1.0 + October 2007 Hotfix
• 3DMark Vantage Professional 1.0.1
• Quake 4 – Patch 1.4.2
• Crysis – Patch 1.2.1 + HardwareOC Crysis Benchmark Tool 1.3.0.0

Some Information About our Methodology

Since we were reviewing entry-level CPUs, we also used an entry-level video card, a GeForce 8500 GT. Because of that, you cannot compare the results published on this review with results published on our other reviews, where we used a different video card.

Even though we had DDR2-1066 memories, we configured them as DDR2-800 units, for two reasons. First, entry-level PCs won’t use DDR2-1066 memories and, second, Athlon X2 can only access memories up to 800 MHz. Also, we made sure that with both AMD and Intel CPUs the memories were running at the same speed (800 MHz) and timings (5-5-5-18). Since the reviewed Athlon X2 4600+ doesn’t suffer from the memory divider problem (see introduction for more details), memories were really running at 800 MHz with this CPU.

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”]

The new PCMark Vantage program 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.

 

Athlon X2 4600+ achieved a higher PCMark Vantage score than its competitors: 6.06% higher than Pentium Dual Core E2180 and 13.54% higher than Celeron E1400. The score achieved by this AMD CPU was similar to the one achieved by Core 2 Duo E4400.

Here the higher L2 memory cache played a major role, with Core 2 Duo E4400 achieving a score 8.99% higher than Pentium Dual Core E2180 and 16.68% higher than Celeron E1400.

Pentium Dual Core E2180 achieved a performance similar to Pentium Dual Core E2160 and a score 7.06% higher than Celeron E1400’s.

 

On TV and Movies score Athlon X2 4600+ was the fastest CPU, being 5.01% faster than Core 2 Duo E4400, 7.31% faster than Pentium Dual Core E2180 and 10.06% faster than Celeron E2160.

Core 2 Duo E4400 achieved a performance similar to Pentium Dual Core E2180, being 3.44% faster than Celeron E1400.

On gaming score Athlon X2 4600+ was again the fastest CPU, being 7.56% faster than Core 2 Duo E4400, 9.35% faster than Pentium Dual Core E2180 and 13.81% faster than Celeron E1400.

Here Core 2 Duo E4400 and Pentium Dual Core E2180 achieved similar performance, with Pentium Dual Core E2180 being 4.08% faster than Celeron E1400.

Here Athlon X2 4600+ was one more time the fastest CPU, being 7.22% faster than Core 2 Duo E4400, 12.03% faster than Pentium Dual Core E2180 and 18.30% faster than Celeron E1400.

The performance here was proportional to the size of the L2 memory cache: Core 2 Duo E4400 was 4.49% faster than Pentium Dual Core E2180 and 10.34% faster than Celeron E1400. Pentium Dual Core E2180 was 5.60% faster than Celeron E1400.

On the Communications suite Athlon X2 4600+ achieved a performance similar to Core 2 Duo E4400, being 11.88% faster than Pentium Dual Core E2180, 14.07% faster than Pentium Dual Core E2160 and 19.89% faster than Celeron E1400.

The performance here was proportional to the size of the L2 memory cache: Core 2 Duo E4400 was 9.64% faster than Pentium Dual Core E2180 and 17.49% faster than Celeron E1400. Pentium Dual Core E2180 was 7.16% faster than Celeron E1400.

And finally on productivity Athlon X2 4600+, Core 2 Duo E4400 and Pentium Dual Core E2180 achieved a similar performance. Athlon X2 4600+ was 25.75% faster than Celeron E1400 and 26.02% faster than Pentium Dual Core E2160 here.

Pentium Dual Core E2180 was 22.63% faster than Celeron E1400 and 22.89% faster than Pentium Dual Core E2160.

Pentium Dual Core E2160 and Celeron E1400 achieved similar results.

[nextpage title=”DivX Encoding”]

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 new SSE4 instruction set – but no CPU included in this review has this instruction set. , so we expect that CPUs with SSE4 support reach a higher performance here. From the CPUs we included in our review only Core 2 Duo E8400 supports SSE4.

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.

Athlon X2 4600+ was the fastest CPU for DivX encoding, being 4.91% faster than Core 2 Duo E4400, 7.89% faster than Pentium Dual Core E2180, 9.21% faster than Celeron E1400 and 15.77% faster than Pentium Dual Core E2160.

As you can see on the above chart, L2 memory cache makes a small difference for DivX encoding. Core 2 Duo E4400 (2 MB L2 cache) was 3.13% faster than Pentium Dual Core E2180 (1 MB L2 cache) and 4.52% faster than Celeron E1400 (512 KB L2 cache). The performance difference between Pentium Dual Core E2180 and Celeron E1400 was of only 1.44% thus we have to consider them as having the same performance level for DivX encoding.

[nextpage title=”MP3 Encoding”]

For MP3 encoding we used LAME 3.9.7 encoder. This is a freeware encoder included in several commercial programs such as Nero. Since LAME is a command-line encoder, we used RazorLame 1.1.5a to have a graphical interface plus a way to measure the time the CPU took to convert an entire audio CD to MP3 format. The CD we chose was Only Dance: 1980-1984. Even though we like New Wave rock the reason we chose this CD in particular was because it has 20 songs for a total of exactly 75:28.40 minutes, i.e., it is a “full” CD. We extracted all audio files to the hard disk drive as Wav files and converted the files located on the hard disk drive and not on the CD, so the optical drive wouldn’t have any interference on the performance achieved.

The results you can find on the chart below and are on seconds, meaning the lower the value, the better.

On MP3 encoding Athlon 64 4600+, Pentium Dual Core E2180 and Celeron E1400 achieved the same performance level. The only significant performance difference was with Pentium Dual Core E2160, due to its lower clock rate, as you can see on the above chart. Pentium Dual Core E2180 was 9.93% faster, Athlon X2 4600+ was 8.90% faster and Celeron E1400 was 7.53% faster than Pentium Dual Core E2160.

Among the Intel CPUs that run at 2 GHz, the only significant difference was between Core 2 Duo E4400 and Celeron E1400, with E4400 being 3.33% faster than E1400.

[nextpage title=”Photoshop CS2″]

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 CS2 there is a methodology created by the folks at GamingHeaven that is very accurate. Their script applies a series of 12 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 give the total time taken to run the 12 filters from GamingHeaven batch. The results below are given in seconds, so the lower the number the better.

On Photoshop Athlon X2 4600+ achieved the lowest performance among the CPUs we tested: Pentium Dual Core E2180 was 12.11% faster, Celeron E1400 was 9.01% faster and Pentium Dual Core E2160 was 4.99% faster.

Core 2 Duo E4400 achieved the same performance level as Pentium Dual Core E2180.

Pentium Dual Core E2180 was 3.41% faster than Celeron E1400 and 7.49% faster than Pentium Dual Core E2160.

[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) to speed up the process.

On Cinebench the winner was Athlon X2 4600+, which achieved a performance similar to Core 2 Duo E4400 and was 7.32% faster than Pentium Dual Core E2180, 17.54% faster than Pentium Dual Core E2160 and 17.61% faster than Celeron E1400.

Celeron E1400 and Pentium Dual Core E2160 achieved the same performance, showing us that its larger cache compensated its lower clock rate  on this program.

Core 2 Duo E4400 was 5.76% faster than Pentium Dual Core E2180 and 15.90% faster than Celeron E1400. Pentium Dual Core E2180 was 9.58% faster than Celeron E1400. All these three CPUs run at the same clock rate, so on this application the larger memory cache makes a huge difference.

[nextpage title=”3DMark06 Professional”]

3DMark06 measures Shader 3.0 (i.e., DirectX 9.0c) performance. We run this software on its default configuration (1280×1024 resolution with no image quality settings enabled) and besides the 3D score given by this program we also compared the results from its internal CPU benchmark.

All CPUs achieved the same 3DMark06 performance, because the video card was the component limiting 3D performance. This indicates that for the savvy user it may be more interesting to have a more expensive video card than a more expensive CPU. We will talk more about this later.

On the internal CPU benchmark conducted by 3DMark06, Athlon X2 4600+ was the fastest CPU among the ones we included in our comparison, being 3.48% faster than Pentium Dual Core E2180, 3.72% faster than Core 2 Duo E4400, 8.44% faster than Celeron E1400 and 17.28% faster than Pentium Dual Core E2160.

Pentium Dual Core E2180 achieved the same performance level as Core 2 Duo E4400 and was 4.80% faster than Celeron E1400 and 13.34% faster than Pentium Dual Core E2160.

It is important to note that a higher CPU score in this program didn’t translate into a higher 3DMark score, since the video card is nowadays t
he component that most affect gaming performance.

[nextpage title=”3DMark Vantage Professional”]

3DMark Vantage is the latest addition to the 3DMark series, measuring Shader 4.0 (i.e., DirectX 10) performance. We ran this program on its “Entry” profile, which is run at 1024×768. We also included the results for the CPU benchmarking done by this program.

On 3DMark Vantage Entry score all CPUs achieved the same performance level, except for Pentium Dual Core E2160, which achieved a score 4.58% lower than the one achieved by Athlon X2 4600+.

On the internal CPU benchmark conducted by 3DMark Vantage, Athlon X2 4600+ was again the fastest CPU among the ones we included in our comparison, being 6.51% faster than Core 2 Duo E4400, 10.70% faster than Pentium Dual Core E2180, 15.02% faster than Celeron E1400 and 20.46% faster than Pentium Dual Core E2160.

The memory cache size made difference here, with Core 2 Duo E4400 (2 MB cache) being 3.93% faster than Pentium Dual Core E2180 (1 MB cache) and 7.99% faster than Celeron E1400 (512 KB cache). Pentium Dual Core E2180 was 3.90% faster than Celeron E1400.

Pentium Dual Core E2180 was also 8.81% faster than Pentium Dual Core E2160 and Celeron E1400 was 4.73% faster than Pentium Dual Core E2160.

It is important to note that a higher CPU score in this program didn’t translate into a higher 3DMark score, since the video card is nowadays the component that most affect gaming performance.

[nextpage title=”Quake 4″]

We upgraded Quake 4 to version 1.4.2 and ran its new multiplayer demo id_perftest at 1024x768x32 with SMP option enable (which allows Quake 4 to recognize and use more than one CPU), under two scenarios: first with image quality settings configured at “low” and then with image quality settings configured at “high.” You can check the results below, given in frames per second.

Here all CPUs achieved exactly the same performance, meaning that on this game the video card was the component limiting the performance, not the CPU. Nowadays for a better gaming performance it is better to have a more expensive video card than a more expensive CPU with an entry-level video card.

[nextpage title=”Crysis”]

Crysis is a very heavy DirectX 10 game. We updated this game to version 1.2.1 and used the HOC Crysis Benchmarking Utility to help us collecting data. We ran the “island” demo at 1024×768 with low image quality settings, selecting “CPU” when the benchmarking utility asked us what we were benchmarking. The results on the chart below are in frames per second.

Here again all CPUs achieved exactly the same performance, meaning that on this game the video card was the component limiting the performance, not the CPU. Nowadays for a better gaming performance it is better to have a more expensive video card than a more expensive CPU with an entry-level video card.

[nextpage title=”Conclusions”]

We came to several interesting conclusions during this review.

Athlon X2 4600+ achieved a higher performance than its main competitors, namely Pentium Dual Core E2180, Pentium Dual Core E2160 and Celeron E1400, in most scenarios. When it wasn’t faster, it achieved the same performance as these other CPUs. The only program where Athlon X2 4600+ was slower than its competitors was on Photoshop CS2. So if you are building an entry-level PC and want the fastest CPU available and want to spend only up to USD 70 we definitely recommend Athlon X2 4600+.

In fact it is interesting to see that AMD has the advantage on low-end CPUs, since on high-end CPUs Intel has the lead.

Another thing we discovered during our review was that for gaming the CPU won’t play a major role like it used to be in the past. This means that if you use a low-end video card (we used a GeForce 8500 GT during our review) the CPU won’t be the component limiting gaming performance: it will be the video card. So if you are building an entry-level PC and at the same time willing to play games, it makes more sense to pick a cheaper CPU and a more expensive video card than the other way around. Of course if you won’t play games it is better to stick with a cheaper video card (and even using on-board video) and invest on a faster (and more expensive) CPU.

As you already know, our secondary goal in this review was to see the impact the L2 memory cache has on computer performance, since we included three Intel CPUs with the same architecture, same internal clock rate (2 GHz), same external clock rate (800 MHz) but different cache sizes (512 KB, 1 MB and 2 MB).

The answer is not so simple: it will depend on the program.

On MP3 encoding and games (read the above explanation on why) we saw no significant performance difference.

On TV and Movies, DivX encoding and Photoshop we saw only a very small performance difference.

On Music and Communications scores from PCMark Vantage we saw a somewhat significant performance difference.

And Productivity score from PCMark Vantage and on Cinebench, on the other hand, the performance difference was huge.

Our conclusion is this. If the difference in price isn’t big enough – as it occurs between Pentium Dual Core E2180 and Celeron E1400 – buy the CPU with the larger cache. But if the price difference is big (for example, Core 2 Duo vs. Pentium Dual Core), buy Core 2 Duo only if you will really use the extra performance. For the average user that is looking for a PC just for doing basic functions like browsing the Internet, using a word processor, checking e-mails, etc, Core 2 Duo is probably overkill. In this case stay with Pentium Dual Core E2180 or, better, Athlon X2 4600+.