And hope people are willing to pay $60 per game on the app store, well known at this point for freemium games and cheap, casual games.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
A9X Die Photo From iPad Pro Reveals 12-Cluster Graphics, No L3 Cache
- Thread starter MacRumors
- Start date
- Sort by reaction score
And hope people are willing to pay $60 per game on the app store, well known at this point for freemium games and cheap, casual games.
There are some pretty nice games out there that are freemium. The need for speed game is a great example.
Same with games like fifa that have been put on iOS
yep.
On this forum is more appropriate to complain about inexistent "lag & stutter" or other clever things like that....
well, considering A9X performances I wouldn't speak about a minimalist approach.
Bloat on OS X ? A real "man CPU" ? ok your agenda and your level of understanding about the matter is now quite clear...Lol, nice baiting...
OS X needs real man CPUs due to its bloat.. not this mobile stuff.
Nope. synthetic benchmarks. It means computational power and even graphics performances.
The A9X is better than a Core M in many area.
No.
The MacBook Core M is more powerful than the iPad Pro.
http://www.pcworld.com/article/3006...he-ipad-pro-really-isnt-as-fast-a-laptop.html
If they wanted to make an ARM Mac they would've done it when they made the MacBook, no point in waiting a year.
In geekbench maybe. Means nothing.
http://www.pcworld.com/article/3006...he-ipad-pro-really-isnt-as-fast-a-laptop.html
Last edited by a moderator:
Why is this the only site that points these things out with benchmarks? And they did answer why Apple does so horribly with physics tests as well.
Wow, so "hard" to believe PCWorld trying to downplay the iPad while stating they use the Surface ...
Cherry picking the "right" test is a good attempt on their part.
the answer is in the name of the website: PCWorld
what would you expect ?
I have never heard of TabletMark before hand. Why isn't that used instead of GeekBench? I do find it interesting that the GeekBench devs believe that future workloads are small loop able things.
I think the physics test is most telling and appreciate FutureMark for sticking to their guns.
That is true but with xcode now you can upload a bitcode instead of the compiled code for ARM so for app developers the transition from x86 to arm would be really easy. Developers already write apps for iOS and OS X and they share big parts of the code between the two platforms, so it wouldn't be an hard transition for them.
I don't think iOS and OS X will merge soon, but I can imagine universal apps in the future. Of course apps will adapt to the device changing the UI for mouse or touch, but they will share most of the code.
I don't think it will happen in 2016 but give them some years and we'll see something
TBH I'm not an huge fan of synthetic benchmarks, because there isn't one that can give us the exact situation, giving all the variables involved ...
Uhhh... why is everyone just ignoring all that massive highly parallel horsepower in the lower left of the chip? That's some intense logic, and it's clearly replicated into many tiles, and if it's not GPU related, than what IS IT? We're talking a third of the die area here, and no one has even a thought?
Possibly something to do with a camera with more megapixels and being able to record and play 4K video format. There's also now the built-in M motion processor.
Really, the probably the most useless test that seemingly relate to NOTHING in the real world (because it blasts past all others in all real software), but hits some kind of design peculiarities... Well good for you for sticking for what's irrelevant.
Are you talking about the physics test? FutureMark said they know exactly why the physics test comes out the way it does and they would prefer not to write code for just 1 vendor to make them look better when none of the other vendors are having the same problem.
As far as TabletMark (different company) is concerned, supposedly its tests are using open source API's and should supposedly represent real world usage. I am not saying GeekBench results are invalid. I just didn't know there was another system benchmark that was cross platform.
By perpetuating essentially a broken down test, because it relates to NOTHING. They are catering to one vendor.
A benchmark which seemingly has no basis in how actual applications work is beyond useless.
I'm not defending Geekbench either, most abstract benchmarks are broken if you ask me.
The best thing is you pick what you need out of a device and look who is best at doing it.
If you spend 99% of your time rendering, it's obviously a very different usage than browsing the net all day long.
Succeeding in one doesn't mean you'll in the other, and maybe it doesn't matter.
So called general benchmarks are really arbitrary in how they weight all the components.
Come on ! iPad Pro is running iOS 9.1, which sucks :S its not smooth or fast right now :S wait for the Last iOS 9 update right before iOS 10 or iOS 10. bad time comparing A9X vs M Core Atm, tbh
Oh wait. I didn't realize the availability of apps is reflective of the capability of the CPU.
The point is that they are much closer in raw power than most people seem to think.
Irrelevant. The A9X will never run OSX. A MacBook will not run iOS. You are fundamentally confused as to what constitutes an Operating System. ARM is not X86.Geekbench Scores:
Last edited:
A MacBook runs iOS just fine. It's in the iPhone Simulator that comes with Xcode. You can even pass around x86 builds of iOS apps for other people to run on their Macs (if they download Xcode).
A Mac won't run user ARM code, but iOS can run on lots of CPU types.
Given the portability of the 64-bit Mach kernel common to both iOS and OS X, I have zero doubt that some engineer inside some Apple secret lab has full OS X running on some arm64 test platform.
Ok, an Engineer, in a lab, on a Xcode simulator? I quote. "A Mac won't run ARM User Code." Tim has no intention of ever letting this unfold to the masses sanctified by Apple. As far as the Kernel, portability does not mean usability .
Last edited: