Robert --

I do not think you have pinpointed the problem. Yes, at any given point in time, the Rosetta or the Docking program will indicate that they are taking 98 or 99 % of your CPU cycles. This is what they are designed to do. They run at the performance level of the idle task on your system. That means that they run at a priority below that of any other task you are running. Most of us require very little of the CPU for the work we are doing on our systems, so the World Community Grid projects can easily show high utilization of your CPU.

The behavior you are describing could be experienced in a memory constrained system due to swapping in and out of virtual memory. If this were the case, you would not only experience problems with the Rosetta code (Human Proteome Folding) and the Docking code (FightAIDS@home), but it would be more of an issue with the Docking code with its higher memory requirements.

I suspect that some other task running on your system is causing this and that by the time you look to see what the cause is, it has subsided (it may even be interupted by your hitting Alt-CTRL-Delete). Usual suspects such as spyware and anti-virus software come to mind. If you are not regularly cleaning spyware off your system, that would be the first thing I would address. Programs such as SpyBot Search and Destroy or Microsoft AntiSpyware are available for free to accomplish this.

The World Community Grid has been running the Rosetta program for over a year now on tens of thousands of PC's around the world. If it were causing problems such as you describe, I think a few other people would have raised this issue long before this.

If you have been away from your system, it would make sense that the specific memory pages you need when you come back are probably swapped out and the drive powered itself down. You will experience a significant delay in getting those pages back into real memory if you have to wait on the hard drive to spin back up to operational speed. I am sure you have multiple applications running on your laptop. Back when I worked for a living, I would have about 4 or 5 host sessions running on VM, Lotus Notes, and a web browser window or 3 all running at the same time (gee, I wonder what employer might see a mix like that -- lol). In any event, the way applications are written today compared to waaay back when I started, there is little concern about memory impact, so even with 1 gig of real memory, you probably are still experiencing memory swapping.

Paging is almost certainly the issue. All grid tasks run at the lowest possible priority, so everything else on the system gets first shot at the CPU. Whatever CPU utilization you see the grid app taking up, that would normally be 'System Idle Process' if you weren't running it.

The grid apps allocate quite a bit of virtual memory (~200MB for HPF and ~300MB for Fight AIDS). On a 512MB system, this is going to immediately cause the OS to start paging. This can cause the system to appear sluggish initially starting a new grid task.

The apps don't use all this memory though and after a little bit of time, large chunks of that are permanently paged out by Windows easing the memory load on the system. At this point, the system usually becomes more responsive again.

Also, Fight AIDS puts significantly more load on your RAM than does HPF.

As for sluggishness when waking up from the screensaver, this is normal for Windows (especially XP) regardless if a grid app is running or not. Windows pages a lot out to disk for performance reasons. When you let the system sit overnight, Windows sees that you aren't using anything and starts paging things out to make room for caches, buffers and freeing up some RAM. When you come back after several hours, Windows has to start paging stuff back in as apps start to be used. You'll obviously notice this more with less RAM, but it happens on all machines.