Hi,
I have been having issues with BI using the Intel hardware acceleration on my current server (i7-8700) - it starts then seems to stop over time. I have been playing with Codeproject AI - I put it in a newer PC - I7-10700 with a Nvidia 1060 6GB gpu.
It occurs to me that if i try moving the BI software entirely to the newer PC with a Nvidia GPU, it might improve the performance.
Soooo with that being said - whats the best way to test it? Can I install the BI demo on the I7-10700, import my settings and give it a try or do some features not work (Direct to disk, etc)?
I also have a I7-14700 desktop which i just bought from Dell & I could potentially try that also - again it feels like a lot of bouncing around.
Whats the best way to give this a try without a bunch of activations?
Thanks,
Rich
Testing BI5 on new hardware?
Re: Testing BI5 on new hardware?
You would have to ask Support what the present limitations of the Demo mode are. I think your plan will work as long as you deactivate the BI5 licence each time you want to try the licence on a different PC. If you don't do that I believe there is a limit, and it will suddenly no longer activate. Not a major problem even then, as you can ask Support to reset that counter if you explain what happened.
As for Intel acceleration, I found it worked with one video driver and failed with the next, so I was swapping drivers to find the working one, and then trying to stop it updating. I found sub-streams worked better for me, so no more Intel acceleration.
As for Intel acceleration, I found it worked with one video driver and failed with the next, so I was swapping drivers to find the working one, and then trying to stop it updating. I found sub-streams worked better for me, so no more Intel acceleration.
Forum Moderator.
Problem ? Ask and we will try to assist, but please check the Help file.
Problem ? Ask and we will try to assist, but please check the Help file.
Re: Testing BI5 on new hardware?
I've had mixed results with hardware comparison testing between my produuction box (i7-4790) and others with more capable processors. The last significant comprison was with an i5-9700 with only four 8MP cameras compared to the 4790 with 22 cameras of various capabilities including three 4K cameras.
CPU runs at 18% on the 4790 with GPU cruising in the 8~10% area (and doesn't support H265 decoding which was handed off to the CPU until Gen6 iCores). A very sweet running box. Substreams, direct to disk, NO AI, nothing fancy. Lots of motion detection for a clips based setup, but very few alerts set. A rock solid system -- until I solo an H265 4K camera, then BI stutters and gags with a nearly unwatchable display. CPU hops up to 45~50% but the box holds it own.
9700? CPU at 15% and GPU at 9% with substreams, direct to disk, same BI version, no motion detection or alerts -- demo mode. Add 10% CPU load when full screen on this box which DOES decode H265 with the GPU staying steady at around 9~10%. This isn't right. Both of those figures should be in single digits.
I've not played the juggling license key game to dig further into any of it. I've been told that the performance parmeters should now be unaltered on the demo version when in the past there were indeed certain restrictions including non-functioning direct to disk capability.
One difference between the two systems that will obviously have some degree of effect on the display adapter efficiency of a demo box is the gigantic Evaluation Version overlay on everything.
No one has been able to explain to me how much demand this particular 'feature' may have on the CPU in demo mode. If my simple little comparison above is any indication, it's A LOT. This would hardly be a fair comparison with another system even with the other system's complete configuration transfered to the test system. It's also a really poor method of encouraging sales if it actually diminishes the capabilities of the software to any observable extent, let alone to the degree my comparisons have loosely indicated.
More CPU and greater GPU should always provide superior hardware performance when properly configured. That doesn't always mean the software will cooperate, at least not before figuring out why not..., which most customers should not be required to do.
I would personally not consider Evaluation Mode an accurate representation of anything more than the available features. All bets are off on video performance simply because of the overlays, in my opinion of course..., and limited comparison experience.
CPU runs at 18% on the 4790 with GPU cruising in the 8~10% area (and doesn't support H265 decoding which was handed off to the CPU until Gen6 iCores). A very sweet running box. Substreams, direct to disk, NO AI, nothing fancy. Lots of motion detection for a clips based setup, but very few alerts set. A rock solid system -- until I solo an H265 4K camera, then BI stutters and gags with a nearly unwatchable display. CPU hops up to 45~50% but the box holds it own.
9700? CPU at 15% and GPU at 9% with substreams, direct to disk, same BI version, no motion detection or alerts -- demo mode. Add 10% CPU load when full screen on this box which DOES decode H265 with the GPU staying steady at around 9~10%. This isn't right. Both of those figures should be in single digits.
I've not played the juggling license key game to dig further into any of it. I've been told that the performance parmeters should now be unaltered on the demo version when in the past there were indeed certain restrictions including non-functioning direct to disk capability.
One difference between the two systems that will obviously have some degree of effect on the display adapter efficiency of a demo box is the gigantic Evaluation Version overlay on everything.
No one has been able to explain to me how much demand this particular 'feature' may have on the CPU in demo mode. If my simple little comparison above is any indication, it's A LOT. This would hardly be a fair comparison with another system even with the other system's complete configuration transfered to the test system. It's also a really poor method of encouraging sales if it actually diminishes the capabilities of the software to any observable extent, let alone to the degree my comparisons have loosely indicated.
More CPU and greater GPU should always provide superior hardware performance when properly configured. That doesn't always mean the software will cooperate, at least not before figuring out why not..., which most customers should not be required to do.
I would personally not consider Evaluation Mode an accurate representation of anything more than the available features. All bets are off on video performance simply because of the overlays, in my opinion of course..., and limited comparison experience.