No, it's a known problem with the CLAM client. For some reason the sync doesn't seem to be completely linear - either the client requests blocks out of order, or the peer sends them out of order - so a set of blocks that don't link to the current chain on disk get stored in memory as orphans. CLAM only tracks 750 orphans by default so once that fills the sync slows to a crawl. I'm sure you're well aware of this issue as it's been discussed several times in this thread.