e6400 slower than fx-55??

[IntelUser2000]

Ha! It seems that your screen name isn't a coincidence...(just j/k)

For BIOS settings:
I loaded "optimized settings," then made the following changes:

Perhaps you can give other information like enable dual core/core multiplexing?? Sorry but voltage settings don't really tell anything here.

Perhaps you can do a Sandra benchmark?? Do a memory bandwidth test with Buffered and Unbuffered result.

For example with Sandra 2007 I get:
Int Buffered: 2968MB/s
FP Buffered: 2983MB/s

Int Unbuffered: 2783MB/s
FP Unbuffered: 2782MB/s
(To get unbuffered results, uncheck the box next to "Enable Buffering/Block-Prefetch Benchmarks, then click on refresh)

 

[hurtstotalktoyou]

Originally posted by: IntelUser2000
Ha! It seems that your screen name isn't a coincidence...(just j/k)

For BIOS settings:
I loaded "optimized settings," then made the following changes:

Perhaps you can give other information like enable dual core/core multiplexing?? Sorry but voltage settings don't really tell anything here.

Perhaps you can do a Sandra benchmark?? Do a memory bandwidth test with Buffered and Unbuffered result.

For example with Sandra 2007 I get:
Int Buffered: 2968MB/s
FP Buffered: 2983MB/s

Int Unbuffered: 2783MB/s
FP Unbuffered: 2782MB/s
(To get unbuffered results, uncheck the box next to "Enable Buffering/Block-Prefetch Benchmarks, then click on refresh)

I don't have any "multiplexing" options in the BIOS, and I don't have Sandra on my PC. Is it freeware? I'd be glad to install it briefly.

You can view the manual for my board here: http://america.giga-byte.com/FileList/Manual/motherboard_manual_ga-965p-s3_e.pdf

 

[IntelUser2000]

I don't have any "multiplexing" options in the BIOS, and I don't have Sandra on my PC. Is it freeware? I'd be glad to install it briefly.

Yes its free
http://www.majorgeeks.com/download.php?det=4664
If link doesn't work go to www.majorgeeks.com and search for Sandra, and select Sandra Lite.

 

[Duvie]

I just did a quick test...

This is done on a different configuration then listed in the link I gave you....

E6400@3.2ghz (800ddr2) Gigabyte S3

I ran test #1 of TMPGenc in that link....

HiDef AVI trailer (2min29sec) Xmen......720x480 convert to mpeg2 with Autmoatic Bitrate (quality 80, hi bitrate 6000, lo bitrate 2000) motion normal, precision 8 bit

2 cores = 59sec
1 core = 108sec

So as you see dual core gave me a 83.0% increase in the same test....

If configured properly and you were using 100% load of both cores then your 60% increase over the sempron should have meant one core of your C2D at 2.9ghz would have been slower then the sempron...That is highly unlikely.

My guess may be your settings are not taxing enough and if you watched task manager for the C2D you likely would not have been 100% a majority of the time...You likely would have been in the 50% range. By default the plus program has multithreaded enabled.

In task manager do you show 2 cpu graphs?

 

[Duvie]

NOw if you have HDD subsystem issues and that is what slowed you down then it would have been obvious because your cpus would not have stayed at 100% load. they would have been waiting for the HDD to get its operations done. Frankly you would have to really have the HDD subsystem really hosed to take that performance hit.

 

[hurtstotalktoyou]

http://img180.imageshack.us/img180/1619/untitledij9.jpg

CPU usage fluctuates between 50-60% per core.

 

[Markfw]

Originally posted by: hurtstotalktoyou
http://img180.imageshack.us/img180/1619/untitledij9.jpg

CPU usage fluctuates between 50-60% per core.

And what is your point ???? That THAT software doesn't use 2 cores ?So what.

 

[myocardia]

Originally posted by: hurtstotalktoyou
I understand that my numbers are unexpected, but what could be interfering? Since all three apps posted minimal performance increases from system to system, I have to assume any problem is with my hardware or driver configurations. But what, exactly?

Assuming you have DMA enabled through Device Manager, it sounds to me like you're having the problems a few other people have experienced with their C2D overclocks: slightly too little vcore. I know it sounds weird, but it isn't all that isolated. People who were giving their C2D's almost enough vcore have reported their systems being considerably slower, until they bumped their vcore one "notch" higher.

 

[hurtstotalktoyou]

Originally posted by: Markfw900

And what is your point ???? That THAT software doesn't use 2 cores ?So what.

It does use two cores.

 

[Duvie]

Originally posted by: hurtstotalktoyou
http://img180.imageshack.us/img180/1619/untitledij9.jpg

CPU usage fluctuates between 50-60% per core.

Well shite....It is not even using both cores...It is using 1 core and 10% of the 2nd if that...

that is because....

"Low resolution DVD....352x240....CBR 1850kbps"? this probably taxes the sempron but wouldn't tax the C2D running stock 1.86ghz....

Ever wonder why they use HiDef codecs now in the test for cpus???? because cpus are so powerful simple conversion to MPEG2 is nothing. Now you are doing a very low resolution and bitrate and it just isn't stressful for most any cpu.


This is exactly why I take offense to you posting your results. they may get misinterpreted. These are very weak test to come up with the conclusion you did.

 

[Duvie]

Originally posted by: hurtstotalktoyou

Originally posted by: Markfw900

And what is your point ???? That THAT software doesn't use 2 cores ?So what.

It does use two cores.

Not really....It shares between 2 cores but the load would barely stress a second core....If you turned off multithreading in the options tab under environmental and cpu tab and reran the test the highest cpu usage you could get is 50% and your time would only increase by 10%.

This is an example of a weak test for a dual core system....

You would need something in the range of 720x480 resolution, bit rate average of 6000 to 2000 like mine. Even then it may only be 80-100% range...

 

[Duvie]

Drop your speed down to stock and rerun the test. I bet maybe your cpu usage would increase but still likely could not hold both cores to 100%....If you set multithreading off and use just 50% or one core of the E6300 at stock 1.86ghz you should get close to the semprons time at 2.3ghz....2 cores if taxing enough should just start pulling away.


I am starting to think this isn't about the system not being set up right as much as it is a lack of understanding of what you were actually comparing and doing it in a way that actually compares them. Otherwise this is like inviting dual cores to a test competition then use apps that are single threaded. Totally skews the results and leaves you with results that are worthless to anyone other then you.

 

[lodmdma]

Originally posted by: AnonymouseUser
Is dvdshrink using both cores?

dunno...

how do i enabled DMA through Device Manager ??

 

[lodmdma]

anyone?