PPD gone for a walk?

Hey guys, this morning I opened BOINC to make sure everything is going smoothly and I noticed that my PPD for the past while has been over 2200 but is suddenly down to about 200 for the last two days. According to my profile at www.worldcommunitygrid.org , there isn't an unusually high number of errored units, just the usual Human Proteome thingy that gives about two errored units per month (Windows 7 64bit). I've attached an image from BOINC manager below.

What's going on? Am I somehow no longer contributing to the project? Have points awarded been cut by a factor of ten (no big deal but would explain things)? I'm SURE my Phenom 2 @ 3.7GHz should get a higher score :rolleyes:

Oh, I also see that BOINC hasn't requested work for three days?

Oh well here's something. Take note of the page number :rolleyes:

You are going to earn some serious points once you get them validated.

I imagine you were running HFCC and/ or FAAH before switching to CMD2?
HFCC and FAAH are zero redundancy validations, meaning that only one copy of the WU gets crunched with an occasional double to make sure everything is good. CMD2 on the other hand requires that two different people crunch the same WU in order top make sure everything is good. Typically WUs with a quorum of 2 validate in 48 hours with a few coming in later (sometimes much later). Combine that with, another assumption, your low cache setting and what you are seeing make perfect sense. You should start to see better PPD very soon ... probably today.

I haven't swapped projects, since day one I've been running all except the beta projects.

Perhaps you just got a stream of CMD2 WUs back to back.
That subproject has very volatile runtimes so if there are a few that take a really long time then that will mess with your DCF so BM will not ask for more work because it thinks everything is going to take a really long time. As long as you are still returning results (sort your results page by the return time) they will eventually validate.