Yes - But nonetheless lenghthening the pipeline does have som disadvantages, especialle in the case of a pipeline stall.
But other than that i think that the other advantages (Cache, improved HT. etc) over the NW will make prescott faster than an equally clocked Northwood - But Intel better improve the branch prediction engine
As long as we dont see another willamette its fine with me, the improved HT should prove usefull in everyday usage and the 3.4-3.6ghz could probly makeup for the added stage. Hopefully at least...
There is one good thing about making the pipelines longer !
You can Clock higher ! and that is what the people and Intel fanboys want ! good times are coming fore them !
It depends on what they're changing (in the pipeline). Lengthening the pipeline in and of itself does not a higher clockspeed make. If the changes are a redistribution of pipeline functions, then it should facilitate clockspeed, but it's to add new functional stages (such was the case with the K8's 2 additional integer pipeline stages), then it really won't help clockspeed at all.
Also, branch mispredict penalties would also depend on what the new stages do (whether they're redistributed or new functions). If they're new functions, and the integer/branch pipeline does not use these, then there is no added branch penalty. The FP pipeline, and FP operations in general, usually have very little branching.
"We are Microsoft, resistance is futile." - Bill Gates, 2015.
You are about to answer a thread that has been inactive for more than 6 months. If you still wish to proceed, please ensure that your posting is original and does not duplicate or overlap any prior responses to this thread.