One of the only places that I'm aware of AGP's bus being too limited in TODAY'S applications is in the instance of real-time video capture and recording.
This is most evident for most people when trying to use a program like Fraps to recording game footage at high resolutions.
As I understand it, AGP has a serious problem sending data 2 directions at the same time, so when fraps is recording at high resolution source (I play all my games at 1600x1200 if I can) there is a huge performance hit because the bus itself becomes saturated.
What I'm looking for, is a comparison of AGP/PCI-E while recording with fraps. I currently have an AGP based system, but the rest of it is pretty much alright for running current games. But I'd like to be able to use fraps while playing, so I'm trying to see if a PCI-E based system will be what I need.
Right now, I drop to 7-10fps when I turn on recording. Even using the option to record at 1/2 resolution, the 1600x1200 source data still has to traverse the bus before being halved, so my performance still suffers.
All of this is information that I've gathered from asking questions, but not based on any real testing, I don't have the facility to do that. It's possibly all wrong. I'm looking for confirmation.
Anyone know of any statistics, comparisons, or benchmarks in this area?
Thanks!
This is most evident for most people when trying to use a program like Fraps to recording game footage at high resolutions.
As I understand it, AGP has a serious problem sending data 2 directions at the same time, so when fraps is recording at high resolution source (I play all my games at 1600x1200 if I can) there is a huge performance hit because the bus itself becomes saturated.
What I'm looking for, is a comparison of AGP/PCI-E while recording with fraps. I currently have an AGP based system, but the rest of it is pretty much alright for running current games. But I'd like to be able to use fraps while playing, so I'm trying to see if a PCI-E based system will be what I need.
Right now, I drop to 7-10fps when I turn on recording. Even using the option to record at 1/2 resolution, the 1600x1200 source data still has to traverse the bus before being halved, so my performance still suffers.
All of this is information that I've gathered from asking questions, but not based on any real testing, I don't have the facility to do that. It's possibly all wrong. I'm looking for confirmation.
Anyone know of any statistics, comparisons, or benchmarks in this area?
Thanks!
Facebook
Twitter