Tweet
I recently picked up a Core 2 Duo laptop. It has the T7200, 2.0GHz Merom chip.
I was wondering how much faster it would be than my Pentium 4 3.06 desktop. I am not disappointed.
In one project I rendered, more of a test project with moving subwindows, moving text, etc the Core 2 did the MPEG- 2 render in 3:54, vs. 6:42 for the P4.
In another "real-life" project with color correction and other filters the Core 2 went 13:10 vs. 36:39 for the P4.
In the test project the Core 2 could preview at 12fps where the P4 couldn't manage 1fps. In the real-life project the Core 2 previewed at 4.7fps while the P4 was at 0.9 fps.
So in conclusion my Core 2 Duo system at 2GHz is generally about two times faster at rendering and 5-10 times faster at previewing than my 3.06GHz P4.
Pretty amazing. Now my laptop is faster than my desktop. Time to dump the desktop for something with some bite.
If you know me you know I just have to analyze the results a bit.
I'm running Vegas 7 and CPU usage is around 70% for previews and 100% for rendering. So the question is why is the Core 2 so much faster, relatively speaking, than the P4 for previewing rather than rendering? Especially considering Core 2 seems to make better use of both cores for rendering rather than previewing?
My theory is that the Core 2 design is better suited to the calls made when previewing rather than rendering. The much more efficient SSE engine in the Core 2 most likely has a lot to do with it.
Also, MPEG-2 software has already been heavily optimized for the P4 so it doesn't take much performance hit when that long pipeline stalls. Or I should say I don't think it stalls much when MPEG-2 rendering. Most of the branch predictor guess are correct with repetitive tasks like rendering.
It's pretty amazing that the Core 2 can beat the P4 (hyperthreaded) while giving up 1000MHz in rendering.
So if you're considering a move to the Core 2 from a P4 I suggest going for it. It is a remarkable chip.
- Mark
I was wondering how much faster it would be than my Pentium 4 3.06 desktop. I am not disappointed.
In one project I rendered, more of a test project with moving subwindows, moving text, etc the Core 2 did the MPEG- 2 render in 3:54, vs. 6:42 for the P4.
In another "real-life" project with color correction and other filters the Core 2 went 13:10 vs. 36:39 for the P4.
In the test project the Core 2 could preview at 12fps where the P4 couldn't manage 1fps. In the real-life project the Core 2 previewed at 4.7fps while the P4 was at 0.9 fps.
So in conclusion my Core 2 Duo system at 2GHz is generally about two times faster at rendering and 5-10 times faster at previewing than my 3.06GHz P4.
Pretty amazing. Now my laptop is faster than my desktop. Time to dump the desktop for something with some bite.
If you know me you know I just have to analyze the results a bit.
I'm running Vegas 7 and CPU usage is around 70% for previews and 100% for rendering. So the question is why is the Core 2 so much faster, relatively speaking, than the P4 for previewing rather than rendering? Especially considering Core 2 seems to make better use of both cores for rendering rather than previewing?
My theory is that the Core 2 design is better suited to the calls made when previewing rather than rendering. The much more efficient SSE engine in the Core 2 most likely has a lot to do with it.
Also, MPEG-2 software has already been heavily optimized for the P4 so it doesn't take much performance hit when that long pipeline stalls. Or I should say I don't think it stalls much when MPEG-2 rendering. Most of the branch predictor guess are correct with repetitive tasks like rendering.
It's pretty amazing that the Core 2 can beat the P4 (hyperthreaded) while giving up 1000MHz in rendering.
So if you're considering a move to the Core 2 from a P4 I suggest going for it. It is a remarkable chip.
- Mark
Comment