Hacker News Reader

An interesting aside from an iPad Pro review noted that Apple's silicon designers are focused on the best design for the purpose of that particular piece of hardware.

I think that's an important insight on how and why Apple is lapping the field on the processor side.

They've hired the best team they could and freed them from all constraints other than physics based ones.

Take the best available node process, sprinkle in any custom algorithm acceleration necessary for voice/ai/camera/ssd throughput, and wattage consumption vs battery, figure out heat dissipation constraints, then max out performance.

No need to create a variety of silicon at slightly different price points- iPhone XR gets the same chip as XS, despite being a cheaper phone. New iPad does get one, since it has to push more pixels and has thermal headroom.

Meanwhile Intel or Qualcomm are creating very artificially specialized hardware at a multitude of price points- while also having to keep off features on their SoC's that don't have broad applicability.

The genius of this move into hardware is paying off more each generation.

> and that they see the investments that they are making in performance will pay off greatly in the coming years, because suddenly our phone will be more powerful than our desktop.

The only way I can see that happening is if people stop upgrading their desktops for years and years AND phones are able to make steady performance improvements AND people don't stop upgrading their phones regularly.

Larger form factors (both desktop and laptop) have physical advantages when it comes to performance (e.g. cooling, more power), and I don't think they can be beaten by a smaller form factor in a fair competition (e.g. chips at similar process nodes).

If mobile computing has reached maturity like desktop computing has, I can see people holding off their mobile upgrades the same as they hold off their desktop upgrades, which would mute the main way mobile can get a performance edge over desktop. The burn-the-ships software update cycle that's more common on mobile may mitigate that, but it's coming to the desktop so it may not be much of a difference in the future.

I agree with all of this. I'm not arguing that laptops and desktops will disappear. But, when people are deciding what to spend money on, I think laptops and desktops will become increasingly marginalized and niche. Also, I think that it very well might be the case that Apple can deliver performance parity compared to larger form factors. The latest iPad Pro is an indication of where iPhones will be in a year. The new iPad rivals the MacBook Pro 2018 in CPU benchmarks [0]. Now that they are designing their own GPUs, I expect the gap to continue to close.

Why would a consumer spend $1500 on a laptop that they use 1-2 hours a day, compared to a phone they use 8 hours a day and is always with them? I'd argue that we are already at the point where most consumers aren't even at the point of considering traditional "computers" in their decisions.

You touch on a good point about software. I think it is the biggest current stumbling block. It is currently the case that desktops and laptops offer significant workflow advantages over mobile devices, even for consumers.

[0] - https://browser.geekbench.com/v4/cpu/10567048

>Why would a consumer spend $1500 on a laptop that they use 1-2 hours a day, compared to a phone they use 8 hours a day and is always with them? I'd argue that we are already at the point where most consumers aren't even at the point of considering traditional "computers" in their decisions.

I think as long as you act primarily as a _consumer_ of other people's content, this is accurate. As soon as you want to _create_ media of your own in any significant way, you then need a keyboard, a larger screen, a decent pointing device, etc. (maybe not all of the above for a particular creative endeavor, but these sorts of things).

The question then is, what percentage of people create enough to need creative devices, and what percentage are happy just being consumers? I think Apple is focusing on the creatives.

> There is no broad use-case advantage that iOS has over Android (ie, I can communicate on both of them, I can get directions on both, etc).

Compliance and security is the magic of iOS. It's pretty trivial to deploy them securely, whereas with Android is a mess that requires lots of third party stuff. Hopefully Chrome will come to the phone someday to address that.

If you really care about device security, it's pretty difficult to not use an iPhone. The only exception that is a clear winner is for field service and other folks who operate in extreme weather conditions, due to iPhone's narrow operating ranges.

> However, what surprises me is that Apple is doubling down on performance. The processing capability of iPhones continues to rise, even when we see that the software that uses that capability has diminishing returns.

Moving the goal posts on performance, beyond what is reasonably necessary, simply incentivizes and challenges developers to start making bigger and better app experiences. This is a passive form of ecosystem insulation that further separates Apple customers from Android customers.

There was a meme a couple years ago, maybe, that on Snapchat you could tell which one of your friends used and iPhone and which ones used Android purely based on the image quality of their photos/videos (I don’t think this is still true). If Apple can silently get devs to start developing apps for the A12, the gap between an Apple experience and an Android experience gets wider and wider IMO.

For historical context, the iOS/Android Snapchat quality thing was because on Android they would take a screenshot of the preview screen of the camera. As of earlier this year, they're actually using the proper API on Android and so things should look better there.

It doesn't particularly speak to the power of the devices, it was more a way to cope with the spectacular fragmentation of Android hardware.

Apple is a bit strange when it comes to creativity tools. They have Final Cut and Logic, both basically major professional production frameworks for film and music, yet they don't have such thing for 3D content, something like a more intuitive Blender, Modo or Cinema 4D, full 3D production suites. Getting into 3D art & design and operating those tools is still quite hard, harder than it should be.

Ben always does a good job in his weekly articles.

I was thinking AR/VR when I read your paragraph 4, and then you said it at the end. So, I completely agree with you. I think most disruption is behind us, but the new wave of Deep Learning enabled computer vision/speech recognition will probably trigger a new wave of use cases and thus new business/services.

I agree. Imagine AR Kit being manyfold powerful and useful not just with games and gimmicks but useful stuff. But only a small percent of hardware, the latest hardware can run those features. That would considerably hinder the adoption rate and the market won't have enough incentive to double down on it. Instead a good proportion of older hardware able to run those new features will suddenly create a whole new market with just a software release. And I do think iPhone is doing lot more computation than people appreciate. A very easy example is that of the Photos app. Face recognition, scene detection, and object detection among many other things happen on device. Comparing that to Google Photos, none of that happens device; all computing is done on cloud. So it gives Apple an opportunity to decentralise all that computation and optimise ,related costs while also upholding its privacy edge over the competing platforms.

Even before we get to ubiquitous AR (capability), I think the on-device AI tools is a big differentiator. Right now, not so much, but I think public opinion on privacy is shifting a bit with what we're seeing about Facebook, etc.

It's a little like a solution looking for a problem, but it seems like a good bet that we will have some of those problems in the future. I guess more like a solution waiting for a problem?

This is what I'm thinking too. It's pretty clear (at least to me) that Apple is trying to make the future of computing look more like iOS, albeit very slowly, instead of making today's tablets look more like OS X, despite everyone saying that this is what they want. This would explain iPads that are faster than today's MacBooks that don't have a proper filesystem yet. It's like they are redoing computing from the ground up, hence the "What's a computer" commercial.

I’m starting to think that iPhone SoC performance gains are just a side effect of them investing heavily into their controlled hardware ecosystem. But the real payoff is in their wearable SoC. The watch is seeing huge performance gains. And the W1 chip in their headphones is similar. Low power, high performance wearable chips are something their competitors are far behind in. This is where Apple are headed.

It’s like building a high performance supercar to learn and grow in order to understand how to bring that technology to a lower powered and more effecient vehicle.

> There is no broad use-case advantage that iOS has over Android

This isn’t really true. At least, once you start using iMessage it’s no longer true.

If you try to switch to Android you will be silently dropped from any iMessage group chats you are in. You will also no longer be able to send high quality media to iPhone users.