So I have this workstation at my work with the following specs: Phenom II X4 945(3ghz), 4GB DDR3 1333, 160GB 7200RPM, 1TB(2x500GB RAID0) ,HD3850 512mb, and RT.X2. Now when I export a 1 min video clipped together with no effects, it takes about 5-10min. According to the users in CS3 it only took a fraction of the time to export a file of the similar length.
So I went searching around when I found PPMB5 and decided to run it total encoding time was 23min on this PC.
I ran PPMB5 at home on both of my PCs
Core2Duo E8400(3ghz), 4GB DDR2 400, 160GB 7200RPM, HD4850 = Total encoding time 27min
Core i5 650(3.2ghz), 4GB DDR3 1333, 250GB 7200RPM, GTS 450(with cuda hack) = Total encoding time 13min
After running the benchmarks I was really confused, I figured my Core2Duo comp wouldn't do too well, but I didn't think my i5 comp would cut the time in half. I don't have any raid setup on my home comps and were designed as all around comp that I could play games on not for video editing.
So with all that said, how should I tackle the problem of the encoding time of the Phenom pc? For some reason regardless if I save to the raid or to the OS drive the time it takes is roughly the same for the disk-test maybe at the most min faster saving to the raid.
If anyone has any ideas on what I could try to do that would be great.
here are the numbers from the output.txt for each pc
Phenom = "755","333","286","199"
Core2Duo = "308","718","591","359"
Core i5 = "217","320","202","17" (software only = "217","320","202","203")
Note: I will be trying to get some sort of Cuda card and ram for the system soon.
Note2: The system originally was a dual 2ghz xeon, but had a motherboard failure and needed the system up and running asap, and the Phenom was the only thing available locally.
There is usually more than one thing going on in a system to cause 'low' results.
The 755 score on the Disk I/O test is far worse than similar machines, who have scores around the 300 mark. This is probably caused by indexing turned on, background processes and lack of tuning. The encoding times are pretty normal for Phenom X4 with only 4 GB of memory. It is the lack of SSE4.1+ support of AMD chips that make them far less suitable for video editing than Intel chips.
A CUDA card will only improve the rendering, but the encoding is still done by the CPU and that is where the weakness of AMD is. Not much you can do about it, other than a new platform.
Your i5 results are very decent for such a CPU.