CyberLink Community Forum
where the experts meet
| Advanced Search >
Integrated HD3000 with Intel Quicksynch .v. GT520M 1Gb
Reply to this topic
vn800rider
Senior Contributor Private Message Location: Darwen, UK Joined: May 15, 2008 04:32 Messages: 1940 Offline
[Post New]
Not wishing to re-open too many debates, but I thought the results of this basic (but real life) rendering test might be of interest to some folk. Since I now have my dinky laptop up and running properly after my travels, I have been doing some real life testing before I start to do some real life editing. (Also Cranston wanted me to do some comparisons between the HD 3000 and the GT520M.)

Hardware/OS
Asus U36S
Windows 7 Professional 64-bit
Intel(R) Core(TM) i5-2430M CPU @ 2.40GHz
4Gb RAM
HD 3000 integrated graphics
GT 520M 1Gb card
PD10 Ultra


Source files
720p NTSC
9Mb Sanyo Xacti HD

clip number = 50 (varied durations)
transition = fade overlap
Total Duration = 10:18 mins


Produce profile
.mp4
standard PD10 menu choice = high quality NTSC


Test 1
Default graphics HD3000
Quick Synch = On
cpu 60-90%
RAM 62%
No 520M activity
render time = 7:03


Test 2
Default graphics HD3000
QS = Off
cpu 96-100%
RAM 62%
No 520M activity
render time = 16:29

Test 3
Default graphics GT520M
Hardware Acceleration = On
cpu 20%
RAM 63%
520M activity = 98%
render time = 14:13

Test 4
Default graphics GT520M
Hardware Acceleration = Off
cpu 96-100%
RAM 62%
No 520M activity
render time = 18:24

There are very minor mediainfo reported differences in the produced file sizes - probably due to variable bit rate encoding differences but so far I can see very little "quality" differences between them on a small screen.

There are no fancy transitions or corrections applied just the raw clips and the overlap fade between them but it provides food for thought?

Cheers
Adrian



Life is really simple, but we insist on making it complicated. (see below)
Confucius
AMD Phenom IIX6 1055T, win10, 5 internal drives, 7 usb drives, struggling power supply.
Reply
[Post New]
Thanks Adrian!

So… If I’m reading this correctly, in terms of “rendering/producing” the SandyBridge+HD3000 (integrated - 128MB) graphics with Quick Sync on, resulted in a render time that is twice as fast, as when bypassing the HD3000, and instead, using the GT520M (discrete - 1GB) graphics with HA on.

Very interesting results.
And certainly good news (for me anyway), as my similar i7-2630QM / 8GB laptop rig, only has the HD3000. So I thought I was really missing out, by not having a discrete 1GB card option. Maybe not huh?

Thanks again for sharing your results.

This message was edited 1 time. Last update was at Apr 11. 2012 23:48

Click here PDtoots for a collection of PowerDirector Tutorials and Tips
Reply
[Post New]
Quote:

...... but so far I can see very little "quality" differences between them on a small screen.



Could you tell me a little bit more about these quality differences ?
What kind of them have you observed ?
Which "method" had better quality in comparison to another ?

it is important for me ....

Reply
vn800rider
Senior Contributor Private Message Location: Darwen, UK Joined: May 15, 2008 04:32 Messages: 1940 Offline
[Post New]
I put the word quality in quotes because it is almost impossible to quantify such a subjective issue. Software such as mediainfo will give some analysis of the internals of the media file, but of greater importance is how it is "seen" by the user, so player, screen, and content all play a part.

Whilst there were differences in the files as reported by mediainfo, I saw no discernible visible difference in the files on a small (up to 32" screen) but that's only my opinion.

I'm afraid I have deleted the test files so I can't relook at them specifically.

Cheers
Adrian Life is really simple, but we insist on making it complicated. (see below)
Confucius
AMD Phenom IIX6 1055T, win10, 5 internal drives, 7 usb drives, struggling power supply.
Reply
[Post New]
I was wondering what were you able to see, especially that I've heard that other peoople had the same feelings.

here is a one of the very interesting link about it (I hope, that I am not against this forum rules)

http://www.anandtech.com/show/4083/the-sandy-bridge-review-intel-core-i7-2600k-i5-2500k-core-i3-2100-tested/9

Looks like it is not only rumour ....
Reply
Reply to this topic
Powered by JForum 2.1.8 © JForum Team