9 comments

[ 4.9 ms ] story [ 25.1 ms ] thread
Can you imagine any other phone warranting this article? "Recently released phone with new parts faster than previously released phone with older parts."
Certainly. Gadget sites have it as headline news every time a new phone is faster than all previous phones. It only happens several times every year. The iPhone 5 will probably be dethroned by the Optimus G in November.

It's big news for gadget sites, but it apparently only warrants 4 upvotes on Hacker News. That feels about right.

it's worth mentioning that a samsung galaxy s3 on jellybean (4.1) is A LOT faster than the the s3 running on 4.03 tested here.
Does this speed still matters for people who aren't gamers ? Isn't the experience already "good enough" ?
It does for the times where the OS is poorly optimised or slows down to a normally crippling pace.
It doesn't matter. All this tells you is that there is no reason to pick any of the current flagship smartphones over another because of their speed.

That's kind of a big deal, given how small the iPhone is and given its above average battery life. Both those things might only be possible because of an underpowered CPU/GPU – but they are apparently not.

It's quite impressive that Apple apparently didn't have to make any really obvious trade offs, it hits all the high notes across the board (even if it isn't necessarily number one in any one category).

Off the top of my head, it might be relevant to face recognition, or the new real-time (?) panorama stitching, or image processing in scanning apps, or advanced paint apps (properly-smearing finger paint?), or HTML5 apps (say, Google Maps), or Garage Band, guitar amp simulators etc ...

Mostly it's stuff that people only do rarely. But the big exception is the browser -- render time is still something we worry about in desktop browsers, so I bet it'll still mean something for mobile browsers for a while.

Gaming and video rendering are another couple that could use more CPU.
Interesting…

"Fastest smartphone in the land (even though older hardware beats it on 1/3-1/2 of the metrics, which are not evenly weighted by any stretch)".

Also, Jellybean has widely been shown to be more performant than ICS on the same hardware (S3 included), so one wonders why ICS was used (after all, software that was released yesterday should probably be compared to software released in July, not December last year, before the hardware came out), for a fair comparison, right?