Paul, looks like Itanium growth hit the ceiling of HP's customer base.
Now, I will take credit for this prediction. While Intel supporters argued that Itanium would take over RISC and AMD supporters argued for Itanium's demise, I was posting (for at least five years) that Itanium would saturate the HP market (by force of the phaseout of PA-RISC) and then Itanium runs out growth. Can you say n-i-c-h-e?
Your numbers lend support to my speculation.