This site is a participant in affiliate programs for Amazon Services LLC, Adorama, and OWC.

What Does Apple Silicon Mean For Photographers?

In Photo Taco Podcast by Jeff Harmon6 Comments

No doubt you have heard that Apple made big news at their annual developer’s conference (WWDC) announcing they are moving from long-term partner Intel to using custom built processors they are calling Apple Silicon. What does this move to Apple Silicon mean for photographers?

Apple switching from Intel to Apple Silicon should mean faster Mac computers with longer battery life. Great news for photographers who use Macs. However, there are likely to be challenges during the transition and photo editing software like Lightroom and Photoshop may suffer.

Faster Mac Computers

Only time will tell, but Mac users should have their hopes and expectations set high when Apple releases their first computers later in 2020 that offer Apple Silicon as a processor option. Apple has shed the Intel and AMD handcuffs that have been holding them back with where they can take the Mac and I expect it to be significant.

Better Heat Management

Have you ever stopped to think about the lack of fans on an iPhone, iPad, or even an iPad Pro? For more than a decade Apple has proven the ability to build custom processors that can be put in tiny consumer electronics devices and still get significant performance without needing fans. If you have used a Mac computer you know very well that fans are a big part of those devices.

Mac computers have been highly criticized for not handling heat very well. Detractors often talk about how Apple seems to choose form over function and produce computers that won’t operate at the full speeds they should be capable of because they want insanely small/thin computers.

It is true that Mac computers tend to throttle the Intel processors that are inside of them because they don’t do a very good job of dissipating that heat. It is also true that the design of the computers being small and thin makes it really tough to design an acceptable solution to the problem.

So should Apple or Intel get the blame here? Mac computers with Intel processors have this problem but Mac devices (iPhone/iPad) running Apple Silicon do not. So it is an Intel problem then right? What about the new XPS 15 9500 laptop from Dell that uses “vapor chamber cooling” technology and run Intel 10th generation processors pretty well?

When Apple releases an Apple Silicon Mac we are going to find out who gets the blame and my bet is it will go Intel’s direction. Apple has proven they can make devices small and thin with great performance with their iPhone and iPad products, why would it be any different when they go their with their Mac computers?

Better Graphics Performance

Like Intel, I think AMD has also been a partner that has constrained what Apple could do with Mac computers. There are use cases where AMD graphics can outperform NVIDIA, but real-world benchmark tests of the Adobe suite of software by my friends at Puget Systems have consistently shown NVIDIA has an advantage.

Apple is not only switching away from Intel with their move to Apple Silicon, they are going to a System on a Chip (SoC) design that will include graphics processing. Apple also won’t need AMD any longer.

I expect GPU performance to increase significantly with Macs running on Apple Silicon. However, this probably won’t be a massive difference for photographers. Lightroom and Photoshop just don’t do a very good job leveraging GPU processing here in mid-2020. Slow progress is being made with nearly every update to Lightroom and Photoshop, so I expect this to become more important in the coming months and years, it just isn’t today.

How Much Faster Will Apple Silicon Macs Be Over Intel Macs?

Early indicators support speculation that Mac computers are about to get a whole lot faster. Developers who have prototype Mac Mini devices to check out how things will work on Apple Silicon are really positive. Software performance tests are not breaking records, but they are at least competitive with Intel computers.

Those results become more impressive when you consider these tests are being run on a two year old processor designed for the iPad Pro. The results become stunning when you understand that the tests are being run in emulation on the device because no native testing tools are available yet.

So how much faster should a photographer expect a new Mac running Apple Silicon to be? Like anyone else who may be talking about this right now, all I can do is speculate. Given the track record Apple has had with their processors in the iPhone and iPad Pro, I am expecting a pretty significant jump in the performance of their own applications and MacOS. I really think something like 100% faster performance might be within reach. That’s right, 2x faster Macs might be possible running Apple Silicon.

Many of you are probably thinking “Apple just take my money!” Let me temper that idea of 2x faster performance with a major caveat. Notice that I said Apple applications and MacOS? The problem with 2x faster as a prediction is that most photographers don’t use applications that come from Apple to process their photos. Most photographers use third party applications like those from Adobe, Capture One, and Skylum. I don’t think we are going to see the full benefit of Apple Silicon with those applications for a while. I’ll explain more on that in a moment.

Still, what would twice as fast mean for photographers who uses Lightroom and Photoshop? I don’t think it means you get your work done in half the time. It means processing intensive functions like import and export in Lightroom Classic would go faster. You probably wouldn’t see half the time as there are other constraints involved like the speed of the disks you are using. Then again, my testing of Lightroom and different disk speeds has shown me that Lightroom Classic isn’t fully utilizing the full performance of the blazing fast storage speeds of current technologies like Thunderbolt 3 or PCI NVMe. Maybe we would see less than half the time on import and export because Apple Silicon brings efficiencies to the entire pipeline.

I also expect that the effects of moving your sliders in Lightroom Classic would get closer to real time updates as well. In Lightroom Classic 9.3, one of the things Adobe did to try and make those sliders have closer to a real time update feel was to skip updating the thumbnail views in the Navigator and the Filmstrip while you are using the sliders. I think that performance improvements a Mac running Apple Silicon could mean this becomes significantly better without Adobe really having to do anything, and can become amazing if Adobe tunes things to the new processing capabilities.

Longer Battery Life

Obviously specific to Apple’s laptop computers, photographers who use Mac laptops should also expect better battery life. That’s right, more performance and better battery life. In fact, if Apple has to choose one over the other, my guess is they may tone down the performance improvement and choose much longer battery life if they have to.

One of the reasons I use a MacBook Pro for my mobile work over a Windows laptop is the battery life Apple laptops provide. Keep in mind that I do 90% of my edition on my custom built Windows desktop computer at home. I use both Mac and PC with my photography and with my day job as an IT professional and there has been such a massive difference in battery life with Apple laptops it justifies the expense for me. I expect that advantage to improve with Apple Silicon. Here is why.

ARM Better than x86 For Power Consumption

I have to get a little more geeky for just a moment here as we talk about battery life. I have already talked about a lot of good reasons Apple is switching from Intel to their own Apple Silicon, but I haven’t talked about ARM vs x86 yet, and we need to here with battery life.

I don’t want to go into massive details on this here, most photographers simply won’t care, but x86 and ARM are something called instruction sets. Think of it like a language operating systems have to use to talk to the processor. Today, when MacOS needs to get some processing done it has to send a message to the Intel processor and x86 is the language that processor speaks. Tomorrow, when MacOS is running and Apple Silicon processor it will talk to it in ARM – a completely different language.

One of the differences between x86 and ARM is the efficiency of that language. ARM is Reduced Instruction Set Computing (RISC) and x86 is Complex Instruction Set Computing (CISC). That is a lot of technical speak to say that ARM as a processing architecture that is more energy efficient.

Smaller Transistor Size Means Better Power Consumption

Another major factor that points to Apple Silicon being much more energy efficient than Intel is the manufacturing process. As transistor size decreases the processors get faster and more energy efficient. The size of the transistors is measured in nanometers and you may have heard processors compared by how small their capacitors are this way.

Somewhere along the way this naonmeter comparison number became less about the actual physical size of the transistors and more about other efficiency gains with the processor, but since marketing had latched on to that nanometer measurement to represent a new generation of processors, the nanometer metric remains what is used today to describe comparisons of manufacturing between AMD, Intel, Qualcomm, and other processor manufacturers today.

For years Intel had been doing a really good job of making consistent improvement with every generation of their processors. Each generation moving to a smaller transistor size making the processor faster and more energy efficient. However, in recent years Intel this has changed with Intel lagging significantly with their manufacturing processes falling well behind others.

Intel processors are holding their own in the performance category for some applications like Adobe After Effects and Photoshop, but they are being outperformed by AMD processors for Lightroom Classic and Premiere Pro. The real story though is that Intel processors require somewhere between 25% and 90% more power than AMD processors. Intel processors are power hungry which makes the battery life of the current MacBook devices from Apple even more incredible.

How Much More Battery Life Will Apple Silicon Macs Have Over Intel Macs?

Just like processing performance, we can look to iPhone and iPad devices as tremendous examples of how Apple has been able to set the bar on battery life for consumer devices. Why wouldn’t we expect something similar with Mac computers running on Apple Silicon?

I wouldn’t be shocked at all for Apple to tell us that they can get somewhere between 50% and 100% more battery life out of Macs running Apple Silicon instead of Intel. Here in 2020 Apple MacBooks get about 11 hours of battery life and a 50% to 100% increase would be about 16 to 20 hours.

Can you imagine a laptop with 20 hours of batter life? Obviously photographers wouldn’t see 20 hours of battery life as running applications like Lightroom and Photoshop are more intensive that the basic usage tests used to find battery life. Still, it seems totally reasonable to expect that a Mac running Apple Silicon may get somewhere between 6 and 8 hours of full performance photo editing.

Transition Challenges

To this point I know I sound like an absolute Apple fanboy. If you have been reading this site or listening to my podcasts long you will know that I am not. I have been doing the majority of my photo processing on a Windows desktop I built back in 2014 and it has served me extremely well. I am going to have a tough decision in 2021 as that is the year I am planning to replace my desktop computer.

To make this seem like less of a fanboy article, let’s get to the negatives that are coming with this transition to Apple Silicon Mac computers.

Massive Undertaking

As Buddy the Elf would say, this change to Apple Silicon is “ginormous!” Apple isn’t new to this game, they have transitioned from one processing architecture to another a few times now, and that will help. Still, this is a really big change and even Apple can’t escape having some bumps over the next couple of years.

Like all modern operating systems, MacOS is extremely complicated. Even without a massive change of moving to a new processing architecture we have the recent MacOS update to Catalina we can look to as an example of things being bumpy.

I know, I know, you are a photography using MacOS Catalina and have had zero problems. Great, no need to send me that email. There have been a lot more photographers who have let me know of their challenges with MacOS Catalina than I have really had in any prior release. It has been bumpy.

You have three levels of changes happening here. A new processing architecture, which requires a new version of MacOS, and requires a new version of your applications. These aren’t just updates here, this is re-compiling those applications to a new instruction set and it is a really big deal. When there is a problem, you have to work through all of those layers of change.

We can actually look to Microsoft to help us see what a large task is in front of Apple. Microsoft has been trying to get Windows built to run on an ARM-based instruction set implemented by Qualcomm processors since 2011. It hasn’t gone very well.

There could be a lot of reasons this hasn’t gone well for Microsoft. It could be that Microsoft hasn’t really applied the resources necessary to make that transition work. It could be it was an idea before it’s time and the processors that ran ARM instructions weren’t powerful enough. No matter the reason, we have an example of a really large company trying to do a similar thing that Apple is doing where it hasn’t gone well.

Don’t get me wrong. I actually expect Apple to pull this off and get things running pretty smoothly so that we have a complete enough transition to Apple Silicon Mac computers that they can be stable and effective for photographers and all other computing use cases. I just expect a few potholes in the road between here and there.

Challenges For Lightroom and Photoshop

In the Keynote where Apple made this announcement they specifically talked about how they are working directly with Adobe to make sure that the Creative Cloud Suite of applications will be ready on day one of their new Mac computers with Apple Silicon. I expect that means it isn’t just being emulated back to x86 using the Rosetta bridge Apple has created. No, I expect that means Adobe had already begun recompiling their applications so that they run natively on Apple Silicon.

Adobe is a massive company with the resources needed to make sure this goes well. Offering a native version of the Creative Cloud Suite is going to be there on day one just like Apple said it would. I have no doubts. It is what happens after that I am worried about.

There are pretty big differences between Windows and MacOS today. MacOS is a lot more similar to Linux than it is to Windows. It is already a pretty good challenge for developers to build applications that work on both platforms. Photographers can look at a company like Skylum for an example.

Skylum started out as a company named MacPhun originally. MacPhun made a photo editing program called Luminar that was available on Mac only. After a few years of ongoing development, and probably some investment funding, the company changed it’s name to something less Mac specific in Skylum and today offers Luminar (and other programs) for both Mac and Windows.

It is going to be harder for a story like that to happen in the future. It is challenging for software developers to create code that works on both MacOS and Windows today, but at least both are using a common processing architecture in the x86 instruction set. Yes, there are some differences between Intel and AMD processors implementing that x86 instruction set, but they aren’t extreme today. At least not as extreme as the difference will be with Apple Silicon implementing the ARM instruction set and Windows still on the x86 instruction set.

To make it worse, things may not get a whole lot better if Microsoft ever manages to get Windows over to the ARM instruction set. Unlike how things have gone with x86 where Intel and AMD are more similar than they are different, processors implementing the ARM instruction set have a lot more differences. Or better put, software has had more need to create code that is specific to a processor rather than writing code to the instruction set.

That is a lot of technical language I don’t expect most photographers to follow or understand. Development teams throughout the world have to get a whole lot better at something called cross-compiling than they are today in order for their applications to work equally well on both Windows and MacOS in the future.

This is big enough it will probably mean a slow down for a company as big as Adobe while they engineer a plan going forward after they work through the effort to get their applications to natively run on Apple Silicon. Software companies smaller than Adobe may have to have their applications run in the Rosetta 2 emulation for a while, and then either stop supporting one of the platforms or not offer full feature parity on both platforms where they release new features to one platform first.

Photographers Should Be Excited For Apple Silicon Macs

Overall, photographers who use Mac should be extremely excited for Apple Silicon Macs. I know it is going to be extremely tempting for photographers to immediately jump on an Apple Silicon Mac when they are offered at the end of 2020. I totally get that. I am so excited about this new release and really hope I can find a way to invest in one of these to test it out.

That said, if your sole source of income comes from your photography and downtime could cost you your business you probably shouldn’t upgrade your current Mac to one with Apple Silicon when they first come out later this year. Just too many possibilities for issues that could impact your business and you should wait for people to get some hands-on experience with them before diving in.

Did this article help you?

Comments

  1. Pingback: What Does Apple Silicon Mean For Photographers? - Apple News

  2. Excellent points! Though I take issue with one:

    “At least not as extreme as the difference will be with Apple Silicon implementing the ARM instruction set and Windows still on the x86 instruction set.”

    Application programmers normally write in programming languages like C++, where no source code is specific to any chip. It would be near impossible to write something as big as Photoshop CC in assembly, which is the only ‘language’ that is chip-specific. Re-compiling an application for x86, or AMD, or RISC, will not be that big a deal from the programmer’s point of view. The _compiler_ does all the work.

    Driver programmers, OS programmers, and application programmers that write assemble to take advantage of chip-specific features are exempted. But we’re talking Photoshop.

    1. Author

      For example, here is a great article going through all of the differences in instruction sets between processors:

      https://www.extremetech.com/extreme/234216-arm-guns-for-high-performance-computing-with-its-new-vector-instruction-set

      In order to use those you have to do very specialized things when you compile. Cross compiling very generic things to be minimally functional works, but each has specialized instructions that mean a lot of work is involved to squeeze the very most performance out of things and there have always been issues where the instruction set target used to develop and test are good only to find crashes and other functional problems when cross compiled to another by means of only using switches in the compiler.

      This used to work well decades ago. Like everything in the world it has become far more specialized in order to make applications work as fast as they do.

      Evidence of this is the entire concept of Java. It was designed specifically to address this problem. Well, it took it a step further with providing portability with compile once and run anywhere by means of a JVM that translates byte code to native instructions. It works, pretty well, but has a massive performance penalty with the overhead of the JIT byte code compiler to the native instructions but also because those native instructions by necessity have to be extremely generic.

      Yep, you can still cross compile code today, but you can’t truly leverage the power of the processor and the specialized instructions sets each offers to significantly improve performance. Major applications do indeed get built specific to the instruction set of the processor.

      That was really doable with AMD and Intel because they are so close. It won’t be that way as Apple continues forward with Apple Silicon and the goal they have always had to make their software and hardware as tied together as they can get it in order to make things perform well. Apple won’t be afraid in the slightest to provide a custom compiler that will only produce code that will rock on their processors and not work anywhere else.

    2. Thanks Jeff, – great article. I’m reading it on the evening Apple just unveiled the One More Thing announcement of the new M1 chip line up. It is also the night before I take my trusty 2015 13” MacBook Pro to the Genius Bar to see if it can be revived or needs to be replaced (all photos are backed up on ext drives and in the cloud!)

      So your common sense advice is both fun to read and potentially very relevant if my old trusty doesn’t come back home.

      I just discovered your site in the past few days; it’s one of the best.

      Rich

Leave a Comment