Apple CPU vs. Intel CPU.. Fight!


#21

Your options are excellent:

  • iPad 2017 for $329 – Ye Olde iPhone 6s internals, still significantly faster than any bullshit hardware Android is shipping with and will be well into 2018, and DIRT cheap for what you get. Arguably the best bang for the buck of any Apple product… maybe… like… ever? Seriously.

  • iPad Pro 2017 for $499+ – very deserving of pro moniker, 120hz scrolling is insanely satisfying, it’s stupidly fast and you can add pencil or keyboard if you’re into that stuff. Plus two sizes, though unless you like Very Large Things I definitely recommend the 10.5". And this time around both sizes have 4GB RAM like God intended.


#22

The 2017 iPad Pro starts at $649.


#23

Note how the option of “Keep the thing you already bought and try and use it” is missing? :) Good old Apple! Really pushing that planned obsolescence.


#24

The A7 in the ipad mini 2 is 64bit and will be supported in iOS11, so it’s not actually obsolete.

The A6X in my pooping ipad4 is 32bit, so it will be. That A7 is 1.8x faster than my old ipad4 too. I do find the ipad4 pretty pokey compared to my iPhone 6s+ and iPad Pro 2016, but if it wasn’t being obsoleted I wouldn’t bother to upgrade.


#25

Do you really think Apple has an issue with planned obsolesce in tablets?

Just curious, as I believe it is well known that tablets and smartphones have accelerated very rapidly over the past 4 years. Not only that, but iPad sales have fallen dramatically, and a well-cited reason is that they are holding up very well over time.

What older phone and tablet do you use to your full satisfaction?


#26

I mean, full satisfaction seems super overloaded. But I use an original ipad mini and a Galaxy S6 and see no reason to spend money on newer versions of either despite being able to do so if I wanted to. They meet my needs for them.


#27

I agree. I think a lot of the reactions to whatever Apple does is emotional.

Regarding planned obsolesce. I can scream foul about my 2 iPads (2nd Generation and 3rd Generation which my daughter and I are still using) with regards to the upcoming 64 bit change and ignore the (about 7 years) long lifespan of service they gave me and is still giving me.

And ignore that 1st generation Samsung Galaxy Tab which I chucked away after 3 months of attempting to get a real use out of back in those days.

PS: My wife still uses an Iphone 4 from 6 years ago as a handy music player that she can easily link with the Apple TV.


#28

Suppose that’s one way to address sagging tablet sales. Well, at least if they were going to do that, they gave people an attractive.y priced option to upgrade to.


#29

The 2nd hand resale market for Apple products is quite healthy, at least in Singapore. And the prices are not bad either. So selling a last generation Apple phone and upgrading to a new model is very viable. And I know people who do just that.

It’s the same for the expensive MacBook pros. A upgrade is typically about 1/2 the retail cost once you sell the older model if you are inclined to do that.

An overlooked fact is that by controlling the price AND building quality hardware. Consumers get to enjoy “upgrade rebates” without ever requiring retailers or the manufacturer to subsidize the upgrade.

Because there are people who trust the hardware enough to buy and use it secondhand, the 2nd hand market ensures less waste overall.

In my opinion, Apple by choosing to focus on quality, adds more to the economy as a whole compared to the race-to-the-bottom android hardware culture.


#30

This “leak” from Apr 27th puts the a11 at almost exactly that…

Who knows if it is real but it seems credible based on my extrapolations from current and previous hardware.


#31

In actuality, either the A10x was really good, or we’re indeed hitting process limits as @jsnell predicted.

A10, iPhone 7 → 3325
A10X, iPad Pro 2017 → 3887 (1.17×)
A11, iPhone 8 → 4061 (1.04×)

Generationally speaking

iPhone 5 731 -
iPhone 6 1418 1.9×
iPhone 6s 2304 1.6×
iPhone 7 3327 1.4×
iPhone 8 4061 1.2×

That’s the slope of the graph I probably should have used, rather than basing my estimate on “increase over tablet version of SoC”.

As a reminder

Geekbench 4 scores are calibrated against a baseline score of 4000 (which is the score of an Intel Core i7-6600U). Higher scores are better, with double the score indicating double the performance.

Here’s all the intel CPUs roughly around iPhone 8 level single core performance


#32

Didn’t I hear someone say battery life is way up? Is that bigger battery or less power usage?


#33

I believe OLED screens are more power efficient.

In general, contrary to the bizarro Android “slow devices are good mmkay” group-think that goes on here, a faster CPU typically does better on battery life because it plows through the work at an accelerated rate and gets back to idle sleep state faster, which is where the real power savings are.

The exception is when you stay in a highly loaded CPU state for a long time (gaming?), then it depends how much the actual power draw is. But no sane person expects devices playing nonstop games to last 8+ hours on battery.


#34

Never thought about it that way, but that makes sense.


#35

Obviously your weird inability to understand that people prioritize things other than single core performance in a device with a wide variety of use cases aside, this is a simplistic view of performance in real world devices. Apple themselves has gone from the A10 having 2 fast 2 efficiency(slow) configurations to 2 fast, 4 efficiency cores in the A11. They weren’t just inspired by you linking to your MOAR CORES picture on your blog. There’s real world advantages to having them. They didn’t become idiots overnight and just add more slow cores for no reason.

The more simplistic, slower cores actually make a big difference to system performance. I don’t know the exact details of the A11 implementation, but I would be shocked if it isn’t very much a reflection of the ARM big.Little stuff.

https://community.arm.com/processors/b/blog/posts/ten-things-to-know-about-big-little

Number 6: Are the power savings available from big.LITTLE significant at the system level?

Saving fifty percent or more of the power of the CPU subsystem is a significant saving at the system level. When combined with DVFS, power gating, clock gating, and retention modes, big.LITTLE plays in important role in the overall power management of a mobile device, and it brings opportunities for future power reduction as software power management policies evolve and work more closely together to manage shut-down, core migration, voltage, and frequency in a coordinated policy. Bottom line, the power reductions are very good now, and they will get even better.

Number 7: Can big.LITTLE save power on high performance tasks too?

High performance applications have periods of lower intensity, for example when waiting for user input or while the GPU is active. During these periods, existing smartphone SoCs downshift to lower DVFS points and/or idle the cores. From the diagram below, we can see that during play, an HD racing game causes the DVFS mechanisms to idle the dual-core Cortex-A9 CPUs almost half the time, while operating below 1GHz over ninety percent of the time. All of these idle periods and low frequency states map well to LITTLE cores and present the opportunity to save energy, even for a high performance workload like the GT Racer HD game.

A white paper with more details:

Obviously faster chips are good and all, but it isn’t a simplistic trade off of speed vs cores vs power utilization on the overall system level. Oh, and price matters to a lot of people too.


#36

I care about how it feels, and to me, my old iPad Air 2 feels nigh identical to my 9.7" iPad Pro even though it benches roughly half as fast.

I do really like the trutone display, though. That, you can notice.


#37

Apple was INSPIRED BY the fact that chips can’t get to 10 gigahertz, aka, basic physics. Everyone eventually goes to more cores because they have more and more transistors due to shrinking process size, but what they don’t have is the ability to make those transistors run at 5, 6, 7, 8 Ghz.

One of our team members at Discourse complained his iPad Air 2 was feeling slow just the other day, which surprised me. The general rule of thumb is that a 20% (1.2×) performance difference is the minimum people can reasonably notice, which I agree with. But the near 2× deficit between an Air 2 and a 2017 iPad Pro is definitely noticeable. I guess if all you ever did was only stream movies, or something, you wouldn’t notice, but browsing today’s JS heavy web? Unlikely.

Also @stusser you forgot 120hz scrolling on the Pro, which ruins you for anything less.


#38

The metadata in these results indeed suggests yet another case of almost no frequency improvements from going to a commodity foundry 10nm process. On the other hand, even 15% is a ton of pure IPC to find at this stage. The improvements are surprisingly even across the sub-tests, which is not what you’d expect from e.g. them just making the design wider. We already know from A10X that the bigger L2 cache isn’t doing much.

So probably best to just wait and see until there’s some more reliable data and analysis on the micro-architecture.

Just a couple of small points, 3325 isn’t the appropriate baseline in this comparison. Something changed in software to significantly speed up a couple of the sub-tests (it was either in iOS 11 or in Geekbench 4.0.1 -> 4.1.1):

https://browser.geekbench.com/v4/cpu/compare/502975?baseline=3543468

You need to compare results using the same software versions. The median iPhone 7 Plus result on Geekbench is 3460, so that seems like the choice that’s easiest to justify. (For the purposes of the above breakdown, I just chose a random 3460 result).

On the other hand, that A11 number might also go up. There’s one huge outlier (the AES test running at a third of its previous speed; https://browser.geekbench.com/v4/cpu/compare/3967901?baseline=3543468). Odds are that Geekbench will hotfix that ASAP.


#39

The 2016 pro doesn’t have the higher refresh rate. I use the ipad for normal ipad stuff, browsing the web is a primary usage. Does it take 1.2s to load a webpage rather than 0.6s? Perhaps. Not noticeable.

Also if the X wasn’t coming out with a fancy new screen, I would feel no particular impulse to upgrade from my 6S+ to an 8+. The old phone is perfectly fast.


#40

I’ve written a bunch of stuff on going iPad-only here.

There isn’t much I can’t do on my iPad that I need a regular computer to do.

  • Well, obvious is anything 3D game related. This isn’t too big a deal since my PlayStation is my preferred device.
  • I use Tableau for data analysis. This isn’t available
  • Oddly, Word and Pages on iOS don’t let you edit styles
  • I use Blender to work on some 3D models.
  • Photoshop stuff is a wash. Affinity Photo and Procreate are both excellent on the iPad. Affinity photo is a true, Photoshop-level program running on the iPad. Using the Pencil with these two apps is amazing. The negative is none of the filters I like to use are obviously available on the iPad.

Outside of my day job. whenever I leave the house I always have my 12.9" iPad with me. My primary, personal, non-consumption tasks are drawing and writing. Uylsses and Procreate are excellent apps. I was also able to run my D&D character at a con, and reference the PDFs of the rules I have.

My new MacBook Pro I got mainly for the heavy things (like Blender) that I just can’t do on iOS.