• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

New Dell Encoding performance against older Gigabyte

Randy1754

Junior Member
I bought a new Dell desktop to use for encoding video (no heckling please). It has a Quad 6600 4GB 800 DDR2 and 500 GB SATA. I have updated the power supply (Hiper 680watt) and video card (Radeon 3850 1GB). The board is based on a P35 chipset.

My existing is a home built using Gigabyte 695-DS3 board with the Quad 6600, 3GB 800 DDR2, SATA drives, Radeon x1300 video card.

The Giga encodes 3 times faster (literally) than the Dell which shocks me. I used SISandra thinking I would find a config issue with the drive but nada.

Any thoughts??
 
Sorry, thought that was enough - like what? Sorry, I seriously don't know what else relates to performance. Thanks.
 
Drivers, OS, program used, background processes, how many drives do you have? It sounds like the Gigabyte has more than one.
 
Thanks
Vista Ultimate SP1 on both; Premiere CS3 with current updates on both; background processes on both are the same i.e. Symantec AV, Windows Defender; no other third party packages. Gigabyte uses GeForce 7600 GT 256 MB; Dell has ATI Radeon 3800 1GB.

Single SATA drives with no RAID configurations on both machines. The Gigabyte runs most everything off the C:\ drive. Premiere on the Gigabyte also has a scratch disk on a separate drive. The Gigabyte C:\ drive is a WD7500AAKS (750GB at 7500). The Dell only has the single 500GB drive it came with and would have to page to the same drive the Premiere process is already using for scratch.

Does this help?
 
It could be too much RAM.

Seriuosly, if you google "vista 4 GB" you'll find some people saying 4 GB runs badly for them even with SP1. Your other PC has 3 GB so it's safe from this.

Try taking out one pair of sticks (remember that usually means slots 1,3 or 2,4) and see if performance increases.

If that's it either run with 2 GB for now or buy 2 x 512.
 
An interesting thought, I wouldn't have gone there. SP1 was supposed to implement the fixes to the earlier problems, Anandtech discuss them in great detail some time back. I will take a look at this approach. Thanks, this helps...
 
I don't know much about Premiere, but I should think video encoding could be fairly disk intensive, depending on the bitrate of the video. So having two disks on the homebrew system could be making a big difference. It may also be worthwhile to examine the Dell settings: in the bad old days, a lot of OEMs neglected to optimise their systems, especially when it comes to i/o. As to the silicon (cpu, graphics, memory), the Dell is equal or superior in pretty much every way... Did you notice differing cpu utilisation on the systems during the encoding?

This is a fascinating real-life comparison, and I'd be grateful if you'd provide updates as you isolate and rectify the problem with the Dell.
 
I'd say:
The gigabyte could have faster harddrives. It likely is faster.
Having a separate drive for the swap file probably helps a lot since encoding is very memory intensive.
Also, Dell's tend to have extra processes running in the background that could cause the computer to run slower.
 
Both good thought trains, I can add a drive to the Dell and test again. I will see if any Dell overhead processes are running I can kill. I did some research on the 4GB Vista suggestion and although fairly recently true, everything I can find in the past two months seems to encourage 4GB as the "sweet spot". I also agree Dell may have done their best at optimization on the P35 chipset. I will let you know how the testing proceeds. Thanks all!!
 
Will do, but in the mean time here is the test I just ran. Both boxes running the same file. I used Performance Monitor and took some system performance snapshots as well. The Gigabyte completed a 3GB file (1080i, 29.97 fps, 2 pass vbr) in 54 minutes; the Dell took 2 hours and 31 minutes. Performance monitor shows both identical on disk I/O (~40 per second) and memory consumption (~ 32%) but CPU was a very different story. Gigabyte shows 2% of processor with the idle process taking the remainder while the Dell is running consistently at 92% with the idle process consuming the rest.

I will look at Process Explorer but can't begin to understand why identical processors would behave so very differently.
 
Before anyone asks...

Ran CPUZ against both and they are identical for voltage, stepping, model, revision (both G0)...
 
Looked at Proc Explorer and all readings are the same (threads, priority, working set, etc.) except cpu. CPU is showing between 63-75% with explorer so the Perfmon is wrong but Dell is still pegged at about 97%. Also updated the Bios and rebooted, but no change.
 
Originally posted by: Randy1754
Looked at Proc Explorer and all readings are the same (threads, priority, working set, etc.) except cpu. CPU is showing between 63-75% with explorer so the Perfmon is wrong but Dell is still pegged at about 97%. Also updated the Bios and rebooted, but no change.

How many services/processes are both running?
 
Back
Top