This system is very odd, it reports 0 for floating point and integer speeds and therefore claims 0 credit, but due to the quorum rule gets granted credit.

http://setiathome.berkeley.edu/show_host_detail.php?hostid=655007

I've seen several of these and even had one of my own boxes do this while running E@H. It's due (I think) to BOINC trying to run benchmarks at a time when the cpu is already loaded with some outside task. Something gives and the benchmarks get recorded as zero.

The main problem is lowering the average claim for all in the quorum. There is no problem with validity of results. The benchmarks can always be fixed by running them manually if you notice the problem. In any case the next auto run of the benchmarks will fix things.

---

Cheers,
Gary.

hmmmm

---

I also had this exact same problem not long ago on my linux machine. See this thread. No one could explain it (lots of responses to my thread, lol) but I notice that your machine is using XP so maybe there is a boinc in general problem here that may be more common then is thought.

I tried manual benchmarks, restarting but did not fix it. I am waiting for cc4.45 for linux before using this machine for boinc again. Actually when I ran the benchmarks the mips results was actually 'nan', but this must be seen as a zero result on the server side of things.

As you say though the wu still validate correctly and the quoram still means you get credit.