.... but.....
After a technical team meeting (AKA an evening down the pub), the resident Sunshiner has asked me to point out two small points (at least I think these were his points, it all got a bit fuzzy towards then end of the meeting):
1/ "The new Snoreacle T3 kit was only announced in September, so it's a bit early to be writing it off." I would agree but suggest that unless Larry's T3 kit shows the same ramp up as hp's Tukzilla kit and IBM's Pee7 then it will probably be the last generic SPARC server kit from Snoreacle, anything following on will be niche storage "appliances" that use SPARC cores. Despite arriving in the middle-to-tail-end of the recession, both Tukzilla and Pee7 seem to be doing quite well. Larry's new T3 has arrived after the recession and so has an easier launch, so any failure to ramp is going to be terminal.
2/ As all but one of the current Tukzilla servers are blades-based, hp can count them twice in IDC figures, being able to count them once as UNIX servers and then again count them towards the blades figure as well." True, but then IBM also has the same option with their Pee7 blades, it's just their UNIX blades are pants compared to their UNIX rack servers, so hp will be laughing all the way to the bank on that one. Meanwhile, the actual number of Itanium blades to x64 is probably so low as to make little difference to the number of units, they're still outselling IBM on x64 blades by a considerable margin.
Discuss!