New iPad Pro faster than MBP's?

MBX

macrumors 68000
Original poster
Sep 14, 2006
1,841
440
It sounds like these new iPad Pro's are beasts in terms of performance, CPU, etc. Is it faster than a MBP, let's say a 13'' 2016 base model?

And if yes why doesn't Apple start putting their own CPU's in laptops too if they're more efficient and faster?
 

gobikerider

Suspended
Apr 15, 2016
2,022
1,476
United States
It sounds like these new iPad Pro's are beasts in terms of performance, CPU, etc. Is it faster than a MBP, let's say a 13'' 2016 base model?

And if yes why doesn't Apple start putting their own CPU's in laptops too if they're more efficient and faster?
Nope not possible CPU performance it may almost be there but it's not quite and gpu is stuck on crap 12 core power vr architecture I did the math and the iris pro in MacBook Pro atleast 60x as fast/more powerful than the iPads graphics. If your curious were I'm getting these numbers go to geekbench browser website look up compute scores for iPad Pro (A9X) and then MacBook Pro 13inch (i5 6630u) that might be wrong, its the i5 with iris graohics tho. So in conclusion CPU speed closing in quick, but gpu has honestly hit a wall. Hopefully it'll catch up when Apple implements its own custom architecture of it. Also keep in mind being iPad and iOS it's entirely different need wise vs MacBook Pro tradition x64 operating system so really the MacBook Pro having a 60x times quicker gpu isn't a big deal, go look up the vids the iPad Pro can render 4K video at lightning speeds well lets just say it's not slow. I can't wait to see how the new iPads improve upon the already solid performance of the A9X but keep in mind the A11X is going to be the HUGE jump everyone is hoping for, since that will be the 7nm or 10nm processor.
 

thunng8

macrumors 6502a
Feb 8, 2006
861
242
Nope not possible CPU performance it may almost be there but it's not quite and gpu is stuck on crap 12 core power vr architecture I did the math and the iris pro in MacBook Pro atleast 60x as fast/more powerful than the iPads graphics. If your curious were I'm getting these numbers go to geekbench browser website look up compute scores for iPad Pro (A9X) and then MacBook Pro 13inch (i5 6630u) that might be wrong, its the i5 with iris graohics tho.
What the heck are you on? Had a quick look at Geekbench compute

a9x: 17417
https://browser.primatelabs.com/v4/compute/582247
13" 2016 macbook pro:28091
https://browser.primatelabs.com/v4/compute/800590

If you look at pure graphics performance - you can look up gfxbench where the a9X is faster than iris graphics in the macbook pro 13" 2016:
a9x:
https://gfxbench.com/device.jsp?ben...e=Apple A9X GPU&did=27138730&D=Apple iPad Pro

13" 2016 macbook pro:
https://gfxbench.com/device.jsp?ben...pple MacBook Pro (Retina, 13-inch, Late 2016)
 
  • Like
Reactions: ABC5S

garyleecn

macrumors 6502a
Jul 25, 2014
841
142
I remember reading somewhere that geek bench is optimized for A-series CPUs, the real world performance is much different
[doublepost=1496729333][/doublepost]
It sounds like these new iPad Pro's are beasts in terms of performance, CPU, etc. Is it faster than a MBP, let's say a 13'' 2016 base model?

And if yes why doesn't Apple start putting their own CPU's in laptops too if they're more efficient and faster?
Because such thing as iMac pro exists. A-series CPU and GPU can never compare with those, and it's way to expensive, difficult and buggy to maintain an OS in two different architectures.
 

thunng8

macrumors 6502a
Feb 8, 2006
861
242
I remember reading somewhere that geek bench is optimized for A-series CPUs, the real world performance is much different
[doublepost=1496729333][/doublepost]
Because such thing as iMac pro exists. A-series CPU and GPU can never compare with those, and it's way to expensive, difficult and buggy to maintain an OS in two different architectures.
Geekbench is supposed to be optimised for all the platform it supports. There was some controversy with Geekbench 3 where too much emphasis was on encryption (Apple Ax chip are very fast at encryption) - but that has largely been solved with Geekbench 4 where encryption only play a small part in the overall results.

ipad pro is very fast for a low powered tablet device but one benchmark does not tell the whole picture

http://www.pcworld.com/article/3006...he-ipad-pro-really-isnt-as-fast-a-laptop.html
you do realise that in the benchmark that article is mostly relying on - the developers of the benchmark have been accused of cheating and colluding with intel?

http://www.pcworld.com/article/3023...ntel-of-cheating-with-sysmark-benchmarks.html

There is a reason why no site ever mentions tabletMark when doing benchmarks - too unreliable. In fact, that article is just about the only time I've ever seen tabletmark mentioned in a benchmark comparison.
 

jmeas

macrumors regular
Jun 30, 2015
239
198
San Jose
There are articles on this forum (from around the time the iPad Pro 12.9 was released) regarding comparing the speed of an iPad to a Macbook. I remember the discussions comparing the IPP vs the 12" Macbook, and I also remember folks saying that it's hard to compare given that the architectures of the CPUs are different.

Either way, iOS is pretty limiting, so I don't think making iPads faster is going to be as useful as either:

1. apps that can take advantage of the power
2. changes to the app store to encourage developers to make apps that can take advantage of the power
 

DaniJoy

macrumors 6502
Nov 19, 2015
394
331
California
Geekbench is supposed to be optimised for all the platform it supports. There was some controversy with Geekbench 3 where too much emphasis was on encryption (Apple Ax chip are very fast at encryption) - but that has largely been solved with Geekbench 4 where encryption only play a small part in the overall results.



you do realise that in the benchmark that article is mostly relying on - the developers of the benchmark have been accused of cheating and colluding with intel?

http://www.pcworld.com/article/3023...ntel-of-cheating-with-sysmark-benchmarks.html

There is a reason why no site ever mentions tabletMark when doing benchmarks - too unreliable. In fact, that article is just about the only time I've ever seen tabletmark mentioned in a benchmark comparison.
Dude I have iPad pro's. They are great for low power or very specifically optimized workflows. But They are not as fast as any high end laptop. The point was take any benchmark with a grain of salt. They all manipulate or make mistakes.
 

masotime

macrumors 68020
Jun 24, 2012
2,351
1,901
San Jose, CA
Either way, iOS is pretty limiting, so I don't think making iPads faster is going to be as useful as either:

1. apps that can take advantage of the power
2. changes to the app store to encourage developers to make apps that can take advantage of the power
This. Essentially:
  1. iOS needs to evolve a sufficiently advanced SDK that allows for apps to have features on par with desktop equivalents.
  2. The App Store needs to evolve to a point where software houses can earn comparable revenue / profit from app sales of fully featured pro apps.
Fortunately, Apple is beginning to recognize this - the many many changes in iOS 11 are a very good step in this direction.
 

DaniJoy

macrumors 6502
Nov 19, 2015
394
331
California
This. Essentially:
  1. iOS needs to evolve a sufficiently advanced SDK that allows for apps to have features on par with desktop equivalents.
  2. The App Store needs to evolve to a point where software houses can earn comparable revenue / profit from app sales of fully featured pro apps.
Fortunately, Apple is beginning to recognize this - the many many changes in iOS 11 are a very good step in this direction.
Yup. In a couple generations or less Apple tablets will have enough power to be very productive. Bring on the Pro apps please. Glad to see Apple is moving in a direction towards content creation. Drag and drop is very innovative, I remember when I first saw it in windows 3.1... magic!
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.