But Vita doesn't have a mechanical drive like the PSP did. Probably a bigger battery.
Could the OLED be more power-efficient than the smaller LCD of the PSP, at least for video playback?
The PSP ( Original "Fat" ) has a 1800mAh Battery. The Vita has 2200mah battery. So technically, 20% better battery.
That Mechanical drive was responsible for a large part of the power drain.
OLED is supposed to be more power-efficient then LCD ( Think it was mostly because of black required no power, where as the back light is always on with LCD = the big power drain ), but also do not forget, that the screen is bigger.
Give or take, having these specs with better then original battery life, is not bad.
It's my understanding that a direct shrink is somewhat easier and cheaper; however if there are advantages for the other alternative it might be worth the additional effort/resources, which shouldn't be all that much if it's more about adding more units here and there.
At this moment, the 45nm process is very mature. In other words, there is a low failure rate. Going to a 28nm process, can bring a big risk, if the failure rate is high ( especially with a 120mm sized SOC ).
But, 45nm has reached its limits. If they stick to a Cortex A9, then the only few ways to upgrade its speed are very limited. And can bring in more drainage of power. For instance, going Quad Core on 45nm, with Quad Core GPU, ... you are increasing the power drain a lot ( assuming they stay at the same frequency ).
Then again, going to 28nm can bring big risks, if the production capacity is not there. Imagine releasing the iPad3(or 2S whatever its called ), and they have a problem with the SOC production. This can limit quantity the release for months ( and drive people to the competition ). See in the past, Intel vs AMD. Or Nivida vs ATI.
I can't know how Apple diverts engineering resources between SoC generations to be honest. It could be one team only or it could be multiple teams assigned to different projects in parallel (whereby the 2nd scenario with their so far yearly execution makes more sense).
The bigger question is, what is behind the fact that Apple has taken a ARM license. There is only one other company out there, that makes there own custom designs, and today we knew the effect from that ( Krait benchmarks are nice ).
Given the rate of development, it makes sense for them to have a 2 team method.
2010: Team 1 developed new SOC & releases.
2011: Team 2 releases upgraded to that SOC.
2012: Team 1 developed new SOC & releases.
Problem is, given the fact that the current information for the A5x SOC, seems like a upgrade, not a new design. Its a little out of character if you have a yearly altering role.
Unless we call the A5 as the new SOC. But what makes that the A4 then?
Two new SOC after each other?
Either way what is "good enough" exactly given how fierce competition will be starting this and the next year?
That is what people overlook. Technology is going forward very fast. Sure, Apple can take its sweet time, and release a slightly upgraded CPU/GPU. And it will sell. But ... the jump expected this year in Technology is more, then it was last year. Everybody stayed at the Cortex A9 design.
Now we have Krait, A15, Quad Cores, ... 32/28nm ... Several more powerful GPU's. Apple's software has been the driving factor behind the iPad's / iPhone's. But take in account that Android also has made major strives into fixing its short comings.
