We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.

[nextpage title=”Introduction”]

We received from Intel a Core 2 Extreme QX6850 engineering sample, which is now the most high-end desktop CPU available on the market today, featuring four cores, running at 3 GHz internally and 1,333 MHz externally – a new clock rate that will be officially introduced tomorrow morning. Since our engineering sample had its clock multiplier unlocked, we were able to increase its clock multiplier from 9x to 10x, simulating a future quad-core CPU running at 3.33 GHz – a “Core 2 Extreme QX 6950”. We compared this new CPU to several other CPUs from Intel, check it out.

Figure 1: Core 2 Extreme QX6850 engineering sample.

Intel will release tomorrow morning some new CPUs using the new 1,333 MHz external bus. This new external bus works at 333 MHz transferring four data chunks per clock cycle, and that is why it is referred as 1,333 MHz (333 MHz x 4). In reality it doesn’t work at 1,333 MHz. CPUs using this new bus will use the number “50” at the end of its model name.

Core 2 Extreme QX6850 works internally at 3 GHz multiplying its 333 MHz external bus by 9 and has two 4 MB L2 memory caches. Its full specs you can see in Figure 2. Keep in mind that Core 2 Extreme QX6800, a quad-core CPU with similar model numbering, has slightly different specs as it works internally at 2.93 GHz (11 x 266 MHz), not 3 GHz like QX6850.

Figure 2: Core 2 Extreme QX6850 specs.

Since our engineering sample had its clock multiplier unlocked, we were able to configure it at 10, making our CPU to run internally at 3.33 GHz, as you can see in Figure 3. We could run flawlessly all our programs under this configuration, so we included the results of our CPU running under this scenario in order to simulate a future “QX6950” CPU. We put this naming inside quotes because this CPU doesn’t exist yet, we don’t know if Intel will actually release it and even if Intel does we can’t assume that this is the name that they will use for this CPU.

Figure 3: “Core 2 Extreme QX6950” specs – our CPU with a 10x clock multiplier.

Because this CPU is based on the new 1,333 MHz FSB, it requires a motherboard supporting this new FSB. So you won’t be able to install this or any other CPU based on the new 1,333 MHz external bus on older socket LGA775 motherboards, meaning that you probably won’t be able to upgrade your CPU with a new one by just replacing the processor, you will probably need to replace your motherboard as well.

Intel P35 and NVIDIA nForce 680i or 650i chipsets are some of the chipsets that support the new 1,333 MHz FSB. In our review we used a MSI P35 Platinum motherboard, which is based on the new Intel P35 chipset, but we faced a problem with this motherboard during our tests that we need to explain.

We were using DDR2-1066/PC2-8500 memories and we tried to keep them always running at 1,066 MHz, however this was only possible when the CPU external clock was of 1,066 MHz. With other clock rates the motherboard didn’t provide a memory clock multiplier that resulted in 1,066 MHz. With the FSB set at 800 MHz – which was necessary by our Pentium 4 3.4 GHz – the maximum clock rate we could set for our memories was also 800 MHz. For Core 2 Duo E6750, Core 2 Extreme QX6850 and our QX6950 simulation we had the option to set our memories at 800 MHz, 1000 MHz, 1110 MHz or 1333 MHz. We tried to keep them at 1110 MHz but the system was unstable, thus we set them at 1000 MHz, 66 MHz below the clock rate they should be running at. This slight difference shouldn’t impact the final results.

In the table below we summarized below all CPUs included in this review with their main specs. We also added a column called “memory clock” for you to know the clock rate our memories where running when we collected data for each CPU – the clock rate below 1,066 MHz was a limitation from the motherboard we were using.

Unfortunately Intel didn’t provide us a Pentium D or a Pentium Extreme Edition samples for reviewing, and AMD seems to be out of samples and new high-end CPUs, because we haven’t received new samples from them for ages. A pity. If you’d like to see a comparison between the reviewed CPUs and Athlon 64 X2 5000+, please read our Core 2 Extreme QX6700 Review.

[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.
 
Hardware Configuration

• Motherboard: MSI P35 Platinum (1.0 BIOS)
• Memory: 2 GB DDR2-1066/PC2-8500 with 5-5-5-15 timings, two Patriot PDC21G8500ELK modules (512 MB each) and two Corsair CM2X512-8500C5 modules (512 MB each).
• Hard Disk Drive: Samsung HD080HJ (SATA-300, 7,200 rpm, 8 MB buffer).
• Video Card: MSI factory-overclocked GeForce 8800 GTS 320 MB (NX8800GTS-T2D320E-HD OC).
• Video resolution: [email protected]
• Power Supply: Antec Neo HE 550.

Software Configuration

• Windows XP Professional installed using NTFS
• Service Pack 2
• DirectX 9.0c

Driver Versions

• NVIDIA video driver version: 158.22
• Intel Inf chipset driver version: 8.3.0.1013
• Audio driver version: Realtek R1.62
• LAN driver version: Realtek 5.664.205.2007

Used Software

• SYSmark2004 – Patch 2
• PCMark05 Professional 1.2.0
• GamingHeaven Photoshop Benchmark V2
• Cinebench 9.5
• 3DMark06 Professional 1.1.0
• Far Cry – Patch 1.4 with HardwareOC Far Cry Benchmark 1.8 utility
• F.E.A.R. – Patch 1.08
• Quake 4 – Patch 1.4.2

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=”SYSmark2004″]

SYSmark2004 is a program that simulates the use of real-world applications. Thus, we consider this the best software to measure, in practical terms, the system performance.

The benchmarks are divided into two groups:

• Internet Content Creation: Simulates the authoring of a website containing text, images, videos and animations. The following programs are used: Adobe After Effects 5.5, Adobe Photoshop 7.01, Adobe Premiere 6.5, Discreet 3ds Max 5.1, Macromedia Dreamweaver MX, Macromedia Flash MX, Microsoft Windows Media Encoder 9, McAfee VirusScan 7.0 and Winzip 8.1.
• Office Productivity: Simulates the use of an office suite, i.e., simulates sending e-mails, word processing, spreadsheets, presentations, etc. The following programs are used: Adobe Acrobat 5.05, Microsoft Office XP SP2, Internet Explorer 6.0 SP1, NaturallySpeaking 6, McAfee VirusScan 7.0 and Winzip 8.1.

This software delivers several results, all of them using a specific SYSmark2004 unit. First we have a SYSmark2004 overall score. Then we have a group result for each batch listed above. And for each batch, we have specific results: 3D Creation, 2D Creation and Web Publication for Internet Content Creation; and Communication, Document Creation and Data Analysis for Office Productivity.

For a better visualization, we separated the results into three graphs: overall score, Internet Content Creation score and Office Productivity score.

Because of its new clock rates Core 2 Extreme QX6850 achieved an Internet Content Creation performance between 3.44% and 11.76% higher than its predecessor, Core 2 Extreme QX6700.

Because of its new clock rates Core 2 Extreme QX6850 achieved an Office Productivity performance between 4.07% and 9.83% higher than its predecessor, Core 2 Extreme QX6700.

[nextpage title=”PCMark05 Professional”]

PCMark05 Professional measures the system performance by running several tests. We selected two batches for our comparisons, System and CPU.

The System batch performs the following tests: HDD XP Startup, Physics and 3D, 2D Transparent Window, 3D Pixel Shader, Web Page Rendering, File Decryption, 2D Graphics Memory – 64 lines, HDD General Usage and three multithreading tests.

The CPU batch performs the following tests: File Compression, File Decompression, File Encryption, File Decryption, Image Decompression, Audio Compression and two multithreading tests.

The results are given in a PCMark05 specific unit.

[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.

[nextpage title=”Cinebench 9.5″]

Cinebench 9.5 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.

This software provides five results, Rendering 1 CPU, which measures the rendering performance using just one CPU, Rendering x CPUs, which measures the rendering performance using all CPUs available on the system, Cinema 4D shading, OpenGL Software Lighting and OpenGL Hardware Lighting. Since we were interested in measuring the rendering performance, we are going to compare the “Rendering x CPUs” results from all CPUs. Keep in mind that even though the Pentium 4 CPU we included in our review has only one core, it has Hyper-Threading technology, which simulates two CPUs.

[nextpage title=”3DMark06 Professional”]

3DMark06 is the latest version of 3DMark franchise, measuring 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.

Here it is worth mentioning that Core 2 Extreme QX6850 achieved the same performance level of QX6700 and “QX6950”.

[nextpage title=”Far Cry”]

Far Cry is a heavy game based on the Shader 3.0 (DirectX 9.0c) programming model. We’ve updated the game to version 1.4. To measure the performance we run four times the demo created by German magazine PC Games Hardware (PCGH) and the results presented below are an arithmetic average of the collected data. We used the HardwareOC Far Cry Benchmark 1.8 utility to help us collecting the data.

We ran this game in two scenarios, both at 1600×1200. The first one, which we called “low”, was with no anti-aliasing, anisotropic filtering set to one and maximum details. The second one, which we called “high”, was with 8x anti-aliasing, 16x anisotropic filtering and ultra details. The results below are given in frames per second.

[nextpage title=”F.E.A.R.”]

F.E.A.R. is a heavy game and we used its internal benchmarking module. We upgraded it to version 1.08 and measured performance in two scenarios, both at 1600×1200 with “computer settings” at “maximum”. The first one, we called “low”, was with “graphics card” set at “low”, and the second one, we called “high”, was with “graphics card” set at “maximum”. Let’s take a look at the results, given in frames per second.

On F.E.A.R. with image quality set to “low” all Core 2 CPUs achieved the same performance level, indicating that the video card is the component limiting the maximum performance you can achieve.

[nextpage title=”Quake 4″]

We upgraded Quake 4 to version 1.4.2 and ran its new multiplayer demo id_perftest at 1600x1200x32 under two scenarios: first with image quality settings configured at “low” and then with image quality settings configured at “high”, both of them with SMP mode turned on through the command r_useSMP 1. You can check the results below, given in frames per second.

On Quake 4 with image quality set at “low” all Core 2 CPUs achieved the same performance level, indicating that the video card is the one limiting the system performance.

When we set image quality to “high” the same thing happened.

[nextpage title=”Overclocking”]

Core 2 Extreme QX6850 works internally at 3 GHz multiplying its external clock by 9 (333 MHz x 9 = 3 GHz).

The engineering sample we’ve got was with its clock multiplier unlocked and we could set it to x10, simulating a “QX6950” CPU, i.e., a Core 2 Extreme running internally at 3.33 GHz and externally at 333 MHz (i.e., 1,333 MHz). The results for this overclocking were added on all tables and charts of our review.

Then we left our sample with its default clock multiplier of x9 and increased the CPU external bus clock rate. We could increase the CPU external bus up to 378 MHz, what made our CPU to run internally at 3.4 GHz, a 13% increase on its internal clock rate.

Figure 4: Core 2 Extreme QX6850 overclocking.

With this overclocking we saw a performance increase of 6.89% on PCMark05.

We could set the external clock above 378 MHz, but the system was unstable. We only consider an overclock to be successful when we are able to run PCMark05 and Quake 4 four times without crashing.

Overclocking success depends a lot on the motherboard and on the CPU used. Even between two identical CPUs one can reach a higher overclocking than the other, especially if they are from different production batches.

We didn’t play with any fancy configuration, like increasing the CPU voltage. Thus with more time and patience you will probably achieve a better overclocking with Core 2 Extreme QX6850 than we did.

[nextpage title=”Conclusions”]

From our tests we can definitely say that Core 2 Extreme QX6850 from Intel is the fastest desktop CPU available on the market today.

Buying this new CPU, however, only makes sense for people really needing the extra performance brought by it, which averages 7% compared to Core 2 Extreme QX6700.

The best result for Core 2 Extreme QX6850 was on Cinebench – i.e., 3D rendering – where it was 15% faster than Core 2 Extreme QX6700. So if you work professionally with 3D rendering buying this new CPU is definitely worthwhile, especially because it is coming with the same price tag as Core 2 Extreme QX6700 – we think 15% more performance for the same price is a great deal.

Intel is setting Core 2 Extreme QX6850’s pricing at USD 999 for distributors in the USA for 1,000 quantities, the same price of Core 2 Extreme QX6700 (which we expect to have its price reduced this week), what is very interesting, especially when we think that Core 2 Extreme QX6800 was quoted at USD 1,199 (under the same conditions) and should be also have its price reduced this week.

This is a good news if you are a hard-core user that was thinking on buying an Intel quad-core CPU. Now you can get a faster CPU for the same price. Or can buy an “older” CPU for their new lower price.

Of course the price range we are talking about is a distant dream for almost all users, and if you are just a regular mortal like everybody else, you will do just fine with a Core 2 Duo CPU.