Not really, but you give very little detail on what exactly you are doing so this answer is merely a guess. You essentially have 3 options in PD available depending on what hardware and video formats you have.

1) use CL SVRT, low load on both the CPU and GPU as all unedited source footage is simply passed through to the end produced file when input and output formats are matched.

2) use GPU for encoding, or hardware acceleration as its called. Typically creates a high load on GPU as it's doing the rendering and a low cpu load as it's primarly just doing I/O

3) use CPU for encoding. Typically creates a high load on cpu as its doing the rendering and I/O and a low GPU load as it's doing very little

Is GPU always faster, not necessarily, a high end multi core cpu can outperform a mid level GPU so it depends what's under the hood.
Is CPU always faster, not necessarily, a high end GPU can easily outperform even a high end cpu so it depends what's under the hood.
During CPU encoding, are the CPU's always pegged at 100. Sometimes no, depends on what's limiting in your system. At times I/O either windows pagefile or PD can limit effective use of the cpu to render.

Do SVRT, GPU and CPU encoding produce the same quailty. Not necessary. depends on hardware and produced formats.

Jeff

Quote: Not really, but you give very little detail on what exactly you are doing so this answer is merely a guess. You essentially have 3 options in PD available depending on what hardware and video formats you have.

1) use CL SVRT, low load on both the CPU and GPU as all unedited source footage is simply passed through to the end produced file when input and output formats are matched.

2) use GPU for encoding, or hardware acceleration as its called. Typically creates a high load on GPU as it's doing the rendering and a low cpu load as it's primarly just doing I/O

3) use CPU for encoding. Typically creates a high load on cpu as its doing the rendering and I/O and a low GPU load as it's doing very little

Is GPU always faster, not necessarily, a high end multi core cpu can outperform a mid level GPU so it depends what's under the hood.
Is CPU always faster, not necessarily, a high end GPU can easily outperform even a high end cpu so it depends what's under the hood.
During CPU encoding, are the CPU's always pegged at 100. Sometimes no, depends on what's limiting in your system. At times I/O either windows pagefile or PD can limit effective use of the cpu to render.

Do SVRT, GPU and CPU encoding produce the same quailty. Not necessary. depends on hardware and produced formats.

Jeff

Thanks for your reply, but i believe both CPU & GPU options are available for use at the same time. I have checked the "Enable OpenCL technology" as well as "Enable hardware decoding", what does that mean ? There is actually no option that tells me to only use CPU or only use GPU, & about SVRT, i can't really use it when producing a .mov QuickTime video, but i want to know how to only use cpu or gpu for the rendering process. I can also upload pictures of the settings i use if needed.

Quote: Thanks for your reply, but i believe both CPU & GPU options are available for use at the same time.

Please, show me how to get simultaneous GPU and CPU rendering.

Quote: There is actually no option that tells me to only use CPU or only use GPU, & about SVRT, i can't really use it when producing a .mov QuickTime video, but i want to know how to only use cpu or gpu for the rendering process. I can also upload pictures of the settings i use if needed.

In the Produce area is a Fast video rendering technology: SVRT and Hardware(GPU) video encoder are options. Not using fast video rendering technology will result in cpu rendering for the entire project. Again when using the technology, any footage that does not fall into SVRT or GPU capable (not all brands of GPU's have the same capability) will by default get cpu rendering.

Jeff

You have a way with words, Jeff! Even I understood that explanation. Thank you.

Cheers - Tony
Visit PDtoots. PowerDirector Tutorials, tips, free resources & more. Subscribe!
Full linked Tutorial Catalog PDtoots happily supports fellow PowerDirector users!

Quote:

Quote: Thanks for your reply, but i believe both CPU & GPU options are available for use at the same time.

Please, show me how to get simultaneous GPU and CPU rendering

Here's a picture, it fully uses 1 of my GPUs (core clock & memory frequencies) & goes up to 50% CPU usage but i guess the slow rendering is caused because producing a .mov video with PD10 doesn't support True Velocity (or hardware acceleration) but instead, it uses both but doesn't take full advantage of any of the CPU or GPU so that's my explanation to the slowness (i might be wrong though).

http://img69.imageshack.us/img69/2221/82486185.png

double post mistake..

I don't think PD really supports crossfire or sli. Been discussed previously in PD8, PD9, and I've seen nothing that it is supported in PD10, a little background here with a back to back test. http://forum.cyberlink.com/forum/posts/list/14356.page

QT mov rendering I think is a little special, what actually is doing the work I believe is the CLQTKernel....process. See it in your task manager at 20% plus another 19% for PD. Your GPU load is 0% on both cards, Nvidia does not do hardware acceleration for mov files (HA should have been grayed out on the PD produce page), you might get a little activity if you had "Enable preview" ticked.

Bottom line, you are only using your cpu for rendering, performance for mov probably depends how multithreaded the client interface to QT, CLQTKernel.... process is for your 6 cores so the answer to your question

Quote: I'm asking if there's anyway i could make PD10 use more of my CPU to render videos faster, is there a way to do that ?

not really is still my answer. Buy more cpu power.

Jeff

Quote: I don't think PD really supports crossfire or sli. Been discussed previously in PD8, PD9, and I've seen nothing that it is supported in PD10, a little background here with a back to back test. http://forum.cyberlink.com/forum/posts/list/14356.page

QT mov rendering I think is a little special, what actually is doing the work I believe is the CLQTKernel....process. See it in your task manager at 20% plus another 19% for PD. Your GPU load is 0% on both cards, Nvidia does not do hardware acceleration for mov files (HA should have been grayed out on the PD produce page), you might get a little activity if you had "Enable preview" ticked.

Bottom line, you are only using your cpu for rendering, performance for mov probably depends how multithreaded the client interface to QT, CLQTKernel.... process is for your 6 cores so the answer to your question

Quote: I'm asking if there's anyway i could make PD10 use more of my CPU to render videos faster, is there a way to do that ?

not really is still my answer. Buy more cpu power.

Jeff

Yup, i guess NVIDIA & QT aren't that friendly,& you are totally right about the CLQTKernel process, it's the QT video producing process & i guess i can't do much about it,& as u know, i've got the world's second best CPU out there (after 3960X) so i can't do much more for faster rendering other than overclocking my CPU maybe, but that's it for this thread, thanks alot for helping .

Quote: Yup, i guess NVIDIA & QT aren't that friendly,& you are totally right about the CLQTKernel process, it's the QT video producing process & i guess i can't do much about it,& as u know, i've got the world's second best CPU out there (after 3960X) so i can't do much more for faster rendering other than overclocking my CPU maybe, but that's it for this thread, thanks alot for helping

I don't know if I'd say they are not friendly, Nvidia does accelerate decoding (playback) and you mentioned you ticked that option in PD. The Nvidia CUDA video encoder is an accelerated library that creates streams complaint with AVC/H.264 (MPEG-4 Part 10 AVC) spec and that's it. It has a narrow focus of video encoding applicability but that's all they advertise for encoding as well. It does what they advertise so that's pretty friendly. You will also see that it does not accelerate encoding of MPEG, WMV...and so on. Basically, it takes advantage of hundreds of CUDA cores to accelerate the encoding of H.264 streams.

That's all I can contribute, so as you said, that's it for this thread.

Jeff