All about CPU
 Latest CPU articles
All CPU articles

Newsletters


  • Ask your question about IT issues
  • Post

Partners

The Games selection

violent : Interactive Buddy Unwind on your interactive buddy: Do anything you want to him, it will earn you money, and you can buy other stuff to torture him with.
violent : More Mindless Violence Basic shooting game, but still so powerful! Use the mouse to take aim and shoot at the little beasties before they get to you. Use Space to reload....
Ads

Sponsored links

Test Setup and Benchmarks

Previous Next
6:00 AM - 11/17/2008 by Patrick Schmid and Achim Roos

System Hardware
CPU I Intel Core 2 Duo E7200 (45nm, 2.53 GHz, 3MB L2 Cache)
CPU II Intel Core 2 Duo E8500 (45nm, 3.16 GHz, 6MB L2 Cache)
Intel Chipset: P45 MSI P45 Diamond, Rev. 1.0
Intel P45, BIOS: 1.4
RAM 4x 1 GB DDR-1066 Crucial BL12864BA1608
HDD Seagate Barracuda 7200.11, 500 GB
7,200 RPM, SATA/300, 32 MB cache
DVD Samsung SH-S183
Graphics Card GeCube Radeon HD 4850
GPU: 625 MHz
RAM: 512 MB GDDR3 (993 MHz)
Power Supply Coolermaster, ATX 2.3, 850 Watt
CPU Cooler Zalman CNPS-9700 LED

System Software & Drivers
Operating System Windows Vista Enterprise Version 6.0 (Build 6000)
DirectX 10 DirectX 10 (Vista default)
DirectX 9 Version: April 2007
ATI Graphics Drivers Radeon 8.7 (Vista)
Intel Chipset INF: 9.0.0.1008
Java Java Runtime Environment 6.0 Update 1

Motherboard: MSI P45 Diamond

Benchmarks and Settings

3D-Games Benchmarks and Settings
Benchmark Details
Crysis Version: 1.2.1
Video Mode: 1680x1050
Overall Quality: low
Demo: CPU-Benchmark2 + Tom’s Hardware Tool
Unreal Tournament 3 Version: 1.2
Video Mode: 1280x1024
Sound and DirectX10; Window off
Video Quality:
Texture Details: 1 Level Details: 1 Demo: vCTF-CONTAINMENT_fly Time: 12/60
World in Conflict Version: 1.0.0.9
Video Mode: 1280x1024
Video Quality: low details
Demo: Game-Benchmark

Audio Benchmarks and Settings
Benchmark Details
iTunes Version: 7.7.0.43 Audio CD (Terminator II SE), 53 min Default format AAC
Lame MP3 Version 3.98 Audio CD "Terminator II SE", 53 min wave to mp3 160 Kbps

Video Benchmarks and Settings
Benchmark Details
Pinnacle Studio 12 Version: 12.0.0.6163
Encoding and Transition Rendering
Private MPEG2-Cam-Movie
Video: 720 x 576 Pixel, PAL, 25 fps, 6000 Kbits/sec
Audio: MPEG Layer 2, 224 Kbits/sec 16 Bit, Stereo 44.1 KHz
File Type: MPEG-2 (DVD Compatible)
TMPEG 4.5 Version: 4.5.1.254
Vieo: Terminator 2 SE DVD (720x576, 16:9) 5 Minutes
Audio: Dolby Digital, 48000 Hz, 6-Kanal, English
Advanced Acoustic Engine MP3 Encoder (160 kbps, 44.1 KHz)
DivX 6.8.3 Version: 6.8.3

- Main Menu -
default

- Codec Menu -
Encoding mode: Insane Quality
Enhanced multithreading
Enabled using SSE4 Quarter-pixel search
- Video Menu -
Quantization: MPEG-2
XviD 1.1.3 Version: 1.1.3

- Other Options / Encoder Menu -
Display encoding status = off
Mainconcept Reference 1.5.1
Reference H.264 Plugin Pro 1.5.1
Version: 1.5.1
MPEG2 to MPEG2 (H.264)
MainConcept H.264/AVC Codec
28 sec HDTV 1920x1080 (MPEG2)
Audio: MPEG2 (44.1 kHz, 2 Channel, 16 Bit, 224 kbps)
Codec: H.264
Mode: PAL (25 FPS)
Profile: Tom’s Hardware Settings for Qct-Core

Adobe Premiere Pro CS3 HDTV
Mainconcept H.264 Plugin 3.2
Windows Media Encoder 9.1 AP HDTV
Windows Audio Encoder 10 Pro
Version: 3.0
NTSC MPEG2-HDTV 1920x1080 (24 sec)
Import: Mainconcept NTSC HDTV 1080i
Export: Adobe Media Encoder

- Video -
Windows Media Video 9 Advanced Profile
Encoding Passes: one
Bitrate Mode: Constant
Frame: 1920x1080
Frame Rate: 29.97
Maximum Bitrate [kbps]: 2000
Image Quality: 50.00

- Audio -
Windows Media Audio 10 Professional
Encoding Passes: one
Bitrate Mode: Constant
Audio Format: 160 kbps, 44.1 kHz, 2 channel 16 bit (A/V) CBR
HD Playback (Blu-ray) PowerDVD 8
Blu-ray - Disc (James Bond - Casino Royale)
Video Mode: 1920x1080p (full screen)
Codec: H.264

Synthetic Benchmarks and Settings
Benchmark Details
3DMark Vantage Version: 1.02
Options: Performance
Graphics Test 1
Graphics Test 2
CPU Test 1
CPU Test 2
SiSoftware Sandra XII SP2 Version 2008.5.14.24 CPU Test = CPU Arithmetic / Multimedia, Memory Test = Bandwidth Benchmark

Talkback
johnbilicki 11/17/2008 12:29 PM
Hide
-1+

Cache is your friend! In example on my moderately modified version of AWStats (web statics programmed in Perl) my socket 754 2.0 GHz 1MB single channel CPU chewed a 120MB access log in 15 seconds flat. My socket 939 2.2GHz 512KB cache CPU took a full minute. Keep in mind the socket 939 processor had a 200Mhz advantage over my socket 754 as well as dual-channel memory support. Both of those benefits had little if any positive effect on the final outcome.

I'd recommend getting the cheapest next increment in cache which would be the E8200. It has 6MB cache instead of 3MB and is $160 which isn't a drastic increase.

Of course more cache doesn't always increase performance. Everything tends to be subjective just like the individual processors.

It would be nice to see an article stressing what types of programs benefit the most from cache. For some people that extra $40 is money well spent while others would see no benefit from spending the extra money.

Slomo4shO 11/17/2008 12:44 PM
Hide
-4+

I guess a 40% increase in power consumption isn't too bad for a 50% increase in clock speeds, although at some point it would be more economical to go with the E8400 or E8500(especially if you live overseas where power costs are almost triple of that in the states).

Thanks for the write-up.

jamesl 11/17/2008 1:21 PM
Hide
-1+

"Of course more cache doesn't always increase performance. Everything tends to be subjective just like the individual processors."

http://www.xbitlabs.com/articles/c [...] html#sect0

ViDER 11/17/2008 1:21 PM
Show
Slomo4shO 11/17/2008 1:39 PM
Hide
-6+

http://www.anandtech.com/casecooli [...] spx?i=3458

I wish you provided data of a overclocked E8500 to 3.8GHz as well in your power consumption charts.

In current market prices, the E7200 is about $70 cheaper than the E8500. So you pay roughly 60% more initially to step up to a E8500.

If we are to assume that you will get at least 2.5 years of use out of the chip and that there is 25 watt power consumption difference between a E8500 @ 3.8GHz and a E7200 @ 3.8 GHz(on average), then if an individual would keep his PC on 24/7 would actually pay an additional $70 in power charges in the 2.5 year span. At this point it would be better go with the E8500 as the extra cache would provide better performance and there is no difference in the overall price of the system over the average lifespan.

I hope to see full power consumption data in future articles.

Homeboy2 11/17/2008 2:47 PM
Show
Homeboy2 11/17/2008 2:50 PM
Hide
-0+

I only paid 150 for my e8500

johnbilicki 11/17/2008 2:53 PM
Hide
-0+

Thanks jamesl, that's a nice comparison across the board. At about 10% performance I'd spend the extra money...not everyone would granted.

Pei-chen 11/17/2008 3:06 PM
Hide
-0+

Did the voltage changed when OCed to 3.4GHz. I know you said no put the power consumption doesn't make sense. 15w more for a 283MHz increase in speed and half the cache of E8500?

Pei-chen 11/17/2008 3:13 PM
Hide
-2+

Did the voltage changed when overclocked to 3.4GHz? I know you said no but the power consumption increase doesn't make sense at both idle and loaded. 10w more at the same clock and half the cache at idle and 15w more with a 233Mhz clock increase and half the cache. It is as if E7200 are inherently less efficient.

Alien_959 11/17/2008 3:18 PM
Show
hairycat101 11/17/2008 3:20 PM
Hide
-7+

I would have been intersted in seeing both the E7200 and the E8XXX clocked at exactly the same speed to for these tests too. That would have shown how much cache is or isn't important for the individual tests performed. This kind of leaves us guessing...

trinix 11/17/2008 3:29 PM
Hide
-3+

@john you are comparing 2 different designs from AMD (who uses another system and has other benefits from cache than intel)and based on that you say everyone should get more cache.

Cache is important, don't get me wrong, but there are some problems with the 8200. The 8200 is harder to overclock to begin with, it's already at 1333 for example.

It really depends on what you need it for and if the money would be wasted or not. The apps described are mostly games and zip. Important for a lot of people, but what difference would it be of use to get a better cpu lesser gpu or the other way around. I think this will help people on budget get a good cpu and gpu.

Cache is still very important for the Core2 design. Now let's see if it's as important for the i7.

trinix 11/17/2008 3:37 PM
Hide
-2+

you can compare the results based on fps too. Crysis needs about 300 mhz more to compensate for the loss of cache.

UT3 is very slightly ahead at 300 mhz advantage

World in Conflict needs more then 300 mhz to compensate for the 3mb loss.

If you follow the link a bit higher, you will find a page describing it will be about 5% overall increase in performance, with games up to 10% increase.

So if you have the money, the premium price for cache is worth it. But if you are on a budget, don't worry about the small loss.

clownbaby 11/17/2008 3:39 PM
Hide
-6+

so, the title should probably be "How To: Get a 3.8ghz Dual Core for $120". Are the titles written before the articles here?

BSMonitor 11/17/2008 3:42 PM
Hide
-0+

Speaking of more E7#00 to come. Newegg already has a 10.5 multiplier version.. E7400 at 2.8GHz for only $145. Bumping FSB just to 1333 already gives you 3.5 GHz!!

Slomo4shO 11/17/2008 3:45 PM
Hide
-1+

Pei-chen :
Did the voltage changed when overclocked to 3.4GHz? I know you said no but the power consumption increase doesn't make sense at both idle and loaded. 10w more at the same clock and half the cache at idle and 15w more with a 233Mhz clock increase and half the cache. It is as if E7200 are inherently less efficient.




Quote :
An increase in processor operating frequency not only increases system performance, but also increases the processor power dissipation. The relationship between frequency and power is generalized in the following equation: P = CFV^2 (where P = power, C = capacitance, V = voltage, F = frequency). From this equation, it is evident that power increases linearly with frequency and with the square of voltage.


I hope this answers your question!

vitreoushumor 11/17/2008 4:13 PM
Hide
-10+

You're using an $80 ($62 at newegg) heatsink to cool your $120 processor so it performs marginally better than a stock clocked $190 processor. Is that really worth $8, especially if most E8500 retail processors can hit 3.8GHz on the stock cooler?

hairycat101 11/17/2008 4:41 PM
Hide
-5+

vitreoushumor :
You're using an $80 ($62 at newegg) heatsink to cool your $120 processor so it performs marginally better than a stock clocked $190 processor. Is that really worth $8, especially if most E8500 retail processors can hit 3.8GHz on the stock cooler?



VERY GOOD POINT. I had a PD805 super overclocked and thought I was cool for while before I did the math and realized I could have had a better CPU/cooler combo for cheaper if I had not had to spend so much extra $$$ on the darn cooler.

x11nt4 11/17/2008 4:49 PM
Hide
-2+

E8500 is $188.99 with free shipping at newegg. This has been their price for at least a week. Not $210+. I don't know what websites you people writing these articles look at, but I always check newegg first for all prices. I'll pay the extra $60 for double the cache.


Comments are closed on this page.

Sponsored links