Looks like it's coming in around the same as a dual G5 @ 1.8GHz which is pretty impressive if you think about it. Quite a shrink!

A15 scores should be *very* interesting.



BTW a current top end MacBook Pro scores around 12000. They're very nice machines indeed

> The iPhone 5 Is More Powerful than the Fastest PowerBook Ever Made

We should not forget that the Geekbench benchmark suite scales very well with multiple cores, and that the iPhone 5 is a dual-core device whereas the PowerBook is single-core. Run a single-threaded application and the iPhone 5 (and any other ARM-based device to date) will still lose.
Apart from that, the PowerBook G4 1.67 GHz scores short of 1000 Geekbench points, which has already been beaten by dual-core 1.0 GHz Cortex-A9 devices released as long as 1.5 years ago. Where was the "curiosity thread" back then? Why now, 1.5 years after the real "news" happened?

http://browser.primatelabs.com/android-benchmarks
http://browser.primatelabs.com/geekbench2/search?q=powerbook+1.67

Sure it is impressive, but other modern smart phones offer similar cpu power. Impressive, but not very surprising IMHO. How old are the G5s today and the Powerbooks (@ AW: I don't need a link with an actual date, it's rather a rhetoric question) .
Btw. I think the new Nokias will offer more innovations and more usability for evryday workload. But of course it depends how good or bad WP8 eventually will be. And if someone reacts allergic to MS: I can generally understand that, but in comparison with Apple I think MS is not worse anymore, and Google is not much better either (i.e. all big three are kind of suspicious).
Anyway, my rather simple S60 based phone does most of the stuff I expect from a mobile phone (calls, sms, gps, organizer, basic cam, browser, email), no urgent need to update.

is very very powerfull look here! :)
http://www.youtube.com/watch?v=NQDQywA42bQ

Andreas_Wolf,
Quote:

As far as I can tell the scaling in some cases is actually pretty abysmal.

>> the Geekbench benchmark suite scales very well with multiple cores

> As far as I can tell the scaling in some cases is actually pretty abysmal.

Some cases? What is with most cases then? ;-)

Power without freedom is fascism.

@Crumb

I actually don't think his point was about using MorphOS on an iPone.


@Fraggle

I actually don't think his point was about using iOS.


@Zylesea

I actually don't think his point was about which smartphone is better.


@Miky060

I actually think I got your point! And I think you should get used to this thought; the Apple HW we are running MorphOS on now are at least 6 years old (and counting) and they still represents the *peak* in PPC desktop/laptop computing. And no, the answer on how to tackle this in the future won't be to rely on some hand-built piece of obscure PPC motherboards, made in batches of #30 units (á la Acube/Aeon), based on some router PPC CPU or whatever! No, the answer can only be to swim *with* the flow of HW evolution (link), not *against* it...



[ Edited by takemehomegrandma 20.09.2012 - 00:30 ]

@takemehomegrandma

God man!
You are a broken record.

Quote:



What do you mean?

> the Apple HW we are running MorphOS on [...] still
> represents the *peak* in PPC desktop/laptop computing

At least regarding desktop computing that would be the G5 Macs, which MorphOS (i.e. the public version of it) doesn't currently support and maybe never will.

https://morph.zone/modules/newbb_plus/viewtopic.php?topic_id=8121&forum=9&start=23 (bottom of topmost posting)

While G5 support would be great, I guess we're sort of lacking a lot of software where it'd really make a difference over a G4.

It would expand the range of equipment people can buy, of course.

> While G5 support would be great, I guess we're sort of lacking a lot of software
> where it'd really make a difference over a G4.

I don't know about "a lot", but I'd definitely know how to utilize the performance of a G5:

https://morph.zone/modules/newbb_plus/viewtopic.php?forum=11&topic_id=7186&start=52

This posting I made more than 2 years ago, so add to that all kinds of HTML5 applications and games that OWB on G4 machines can execute but run too slow.

As Andreas points out, 1080p video is beyond our systems' capabilities (although I have played back many 720p videos with a Quicksilver with a third party CPU card and my current 1.42 GHz FW800).

Also, I frequently bog my system down with multiple browser tabs and concurrent processes.

I don't think you can ever have too much CPU power.

@ takemehomegrandma - What I mean is you keep repeating this call for an ISA change.

"broken record" - you skip

takemehomegrandma,
Quote:

Yes
Yes
Yes
Yes
..and Yes! Mine was just a curiosity, something I read and wanted to share with you here. It's not MorphOS related and yes, maybe I discovered that modern mobile phones beat our hardware only now Andreas.

> yes, maybe I discovered that modern mobile phones beat our hardware
> only now Andreas.

Not only you but apparently also the author of the article you linked to ;-) And I think it's important to relate that to Geekbench -- as I said, run a single-threaded application and the result will look differently.

Quote:

Actually a major chunk of Geekbench is single threaded - that's why it scales so badly.

It looks like around 50%* is single threaded so Amdahl's law working backwards would give a score of 1200 for a single core A6 - still above the G4.


*A Galaxy S3 gives 1200 with 2 cores and about 1800 with 4 cores (from a post on beyond 3D forum).

> Actually a major chunk of Geekbench is single threaded - that's why it scales so badly.
> It looks like around 50%* is single threaded so Amdahl's law working backwards
> would give a score of 1200 for a single core A6 - still above the G4.
> *A Galaxy S3 gives 1200 with 2 cores and about 1800 with 4 cores (from a post on
> beyond 3D forum).

It may scale better from 1 to 2 cores than it scales from 2 to 4 cores. So I don't think we can conclude single-threaded A6 figures from your numbers.