Is iPad Pro 2018 faster than 92% of portable PCs?

[mi7chy]

Someone tweeted that the new iPad Pro 2018 with A12X SoC and 6GB DRAM looped through 1000 iterations of base64 encoding and decoding a 3.6MB jpg file in ~45 secs.

https://twitter.com/viticci/status/1061443135975407616

Also, Apple claim that the iPad Pro 2018 is faster than 92% of portable PCs sold in the last twelve months but that is sort of vague so lets make it easy and compare it with a ten year old Intel Core 2 Duo P8400 2.26GHz laptop with 4GB DRAM. The laptop should lose spectacularly, right?

iOS 12 Shortcuts batch translated to bash for MacOS, Linux, etc.

time (for i in {1..1000}; do base64 test.jpg > /dev/null; base64 -d encoded.jpg > /dev/null; done)

But more likely it's only doing base64 encoding only which is equivalent to:

time (for i in {1..1000}; do base64 test.jpg > /dev/null; done)

3.6MB jpg test file (renamed to test.jpg):
https://www.camerafv5.com/files/photos/raw1/calella2.jpg

Completed 1000 loop of both base64 encoding and decoding in 52 secs (1st result) while base64 encoding only is 17s (2nd result).

Only potential issue with the original iOS Shortcuts batch test is in the second part where it's base64 decoding an unencoded file that will likely result in an 'invalid file' so it's not doing any work. If you notice in the screenshot the test.jpg file becomes larger when base64 encoded to encoded.jpg file.

This is where we need someone with the new iPad Pro 2018 6GB DRAM (4GB DRAM crashes according to tweet) to recreate the Shortcuts batch but only time the base64 decoding part to make sure it's not near zero meaning it's not doing anything.

So, best case the new iPad Pro 2018 is slightly faster than a ten year old laptop and worst case, if the original testing is flawed, it's less than half as fast (compared to 2nd result in screenshot). Run the bash command on your MacBook and/or iMac and share your results.

 

[tromboneaholic]

Why don't you just run Geekbench on your old laptop and share your score?

 

[mi7chy]

Why don't you just run Geekbench on your old laptop?

Real world workload is better than useless synthetic benchmark and has always been the case for decades.

 

[tromboneaholic]

Real world workload is better than useless synthetic benchmark and has always been the case for decades.

But your test isn't representative of my real world usage.

 

[kasakka]

I don't feel Geekbench scores are useful either because they don't represent real world work loads but at the same time the one in OP might not be a good one either. Different CPUs and GPUs will usually excel in different workloads based on their architecture and optimizations.

Even if the iPad Pro is as fast as a good laptop, there isn't really any iOS software to make good use of that power.

 

[tromboneaholic]

I don't feel Geekbench scores are useful either because they don't represent real world work loads but at the same time the one in OP might not be a good one either.

I doubt OP's 10 year old laptop would do a good job editing and encoding 4K video compared to an iPad.

I agree it all depends on what your workload is.

 

[Pressure]

What are you trying to test specifically?

Right now you are simply showing 1 data point to prove something that fits your message.

You are welcome to post your SPEC2006 data points from your 10 year old computer.

 

[Koh Phi Phi]

Sheer speed performance in the iPadPro is kind of useless, since the type of tasks any normal user can perform on iOS (and iOS's Apps) will hardly ever stress the CPU.

The iPadPro's hardware is miles ahead of its software, thus rendering speed stress tests pointless.

 

[profets]

Also, FWIW, the original person who did that test had screen recording running during the entire thing.

 

[Dammit Cubs]

Technically, the iPad did beat your laptop, apple didn’t say by how much...

 

[maxgts]

Powerful enough to run macOS, but we're stuck with iOS.

 

[Aston441]

Powerful enough to run macOS, but we're stuck with iOS.

iOS is actually great because there's far more useful third party software for it today than MaOS.


Whether it's staying in touch or shopping or running all the smart stuff in my house.

All it needs is a trackpad option and the lock taken off the file system.

Then it's a real personal computer.

Other refinements can be added later.

 

[Altis]

Apple has been forever telling us that the tech specs don't matter so much as the user experience.

I don't care how fast the hardware is when the user experience is slow. "Animation" delay (ie. input blocking) everywhere, app/tab reloads constantly, cumbersome UI... it doesn't feel fast when I need to attach more than 5 photos to an email, that's for sure.

 

[Koh Phi Phi]

iOS is actually great because there's far more useful third party software for it today than MaOS. It needs is a trackpad option and the lock taken off the file system.

Then it's a real personal computer.

Other refinements can be added later.

Agreed.

iOS is actually great, and far better suited for a touch environment than MacOS. It simply needs a few limitations to be lifted and a bit more flexibility to be added, and voila! You have true laptop-replacement material right there.

 

[alecgold]

Hmm, I do hope iOS 13 will give us a Good iPad Pro centric update, so we can use all this power and also work more efficiently.
[doublepost=1542043520][/doublepost]Then again, I can do a lot with documents by readdle, it’s just not as efficient at certain tasks as macOS. E.g. sorting and filing large number of files etc.

 

[bjet767]

So what!

My iPad Pro 10.5" is fast, but it runs iOS and that is the limiting factor for its use.

I use Numbers, Pages, Adobe and a few other high end productive programs on both the iPad and MacBook Pro. To summarize it, the iPad API compared to OSX is a pain to use for these types of apps.

I've written it before, as soon as iOS can truly multi-task views, like OSX, Windows, Linux, etc. it will remain primarily a content based device. Watch videos, read, email, social networking and single, non-keyboard based, app use. And if you think I'm bashing iPad think again. I often have far more screen time with my iPad and iPhone than the MacBook.

Now if only Apple would add a touch screen to its OSX line. One can wish!!

 

[NT1440]

What kind of usage is this even supposed to be testing?? I don't get it

 

[tromboneaholic]

What kind of usage is this even supposed to be testing?? I don't get it

Imprecise usage of disparate methods to acheive incomplete data analysis.

 

[darkarn]

Even if so, this is due to the iPad Pro 2018 being the first 7nm device out there.

Give Intel and AMD a year or two, we will be seeing laptops with twice the performance of the iPad Pro 2018

 

[tromboneaholic]

Technically, the iPad did beat your laptop, apple didn’t say by how much...

Well it's more than 1,000 times faster than his own laptop.

If you actually read through his post, you see:
The iPad encoded and decoded a 3.6MB jpg file 1,000 times in ~45 seconds.
His old computer encoded and decoded the file 1 time and then didn't do anything 999 times in ~52 seconds.

 

[mi7chy]

iOS is actually great because there's far more useful third party software for it today than MaOS.

MacOS can already run iOS apps via Xcode. Makes much more sense to run the much more capable and robust MacOS on iPad since you get the best of both worlds with professional MacOS software and mobile iOS apps then add touch capability since it's not impossible for a company like Apple with all of their resources when enthusiasts have touch working on Surface Pro hackintoshes. Much like ChromeOS approach with running a desktop class OS that can also run Android mobile apps.

 

[tromboneaholic]

Even if so, this is due to the iPad Pro 2018 being the first 7nm device out there.

Give Intel and AMD a year or two, we will be seeing laptops with twice the performance of the iPad Pro 2018

We'll be seeing iPads with twice the performance in a couple of years too.

 

[NT1440]

MacOS can already run iOS apps via Xcode. Makes much more sense to run the much more capable and robust MacOS on iPad since you get the best of both worlds with professional MacOS software and mobile iOS apps then add touch capability since it's not impossible for a company like Apple with all of their resources when enthusiasts have touch working on Surface Pro hackintoshes. Much like ChromeOS approach with running a desktop class OS that can also run Android mobile apps.

I don’t know why people like you just can’t accept that macOS is simply never coming to the iPad.

Of course Apple CAN do it, but there’s clearly a different path forward.

 

[mi7chy]

Well it's more than 1,000 times faster than his own laptop.

If you actually read through his post, you see:
The iPad encoded and decoded a 3.6MB jpg file 1,000 times in ~45 seconds.
His old computer encoded and decoded the file 1 time and then didn't do anything 999 times in ~52 seconds.

This is actually false. The bash command is doing the exact same thing as iOS Shortcuts but without the GUI front end. Suggest spending a few minutes learning bash 101 on MacOS, Linux, BSD, etc. Or, simply change the loop to 'for i in {1..1}' and see the difference for yourself.

 

[NT1440]

This is actually false. The bash command is doing the exact same thing as iOS Shortcuts but without the GUI front end. Suggest spending a few minutes learning bash 101 on MacOS, Linux, BSD, etc. Or, simply change the loop to 'for i in {1..1}' and see the difference for yourself.

I’m just trying to understand what workflow you’re trying to simulate here. What exactly would this apply to in real world usage? I don’t use bash or anything so I don’t know what you’re showing with this thread?