Unity3D for ARM Macs

So recently, there have been multiple rumors that suggest Apple will announce their transition to ARM Macs this WWDC. Does anyone here know if Unity is planning to support this anytime soon?

I'm planning to upgrade my Mac within the next 2 years and I'm kinda thinking if I should upgrade sooner and get an Intel Mac if Unity will take years to support ARM Macs. My worry is that Blender and Unity will take a long time to support these ARM Macs.

 

We will need to wait to hear what details come out of WWDC before anyone can answer your question. Unity may or may not have inside info, but I wouldn't be surprised if they didn't. In that case, of course they will be interested, but a game engine editor has got to be one of the hardest applications to rewrite for RISC.

 

Yeah. We'll probably not hear anything until after the announcement (if it even happens). But I'm just wondering how ready Unity is for such a shift. I know Windows has been toying with a shift to ARM for years and it's still a mess. But knowing Apple, I'm doubtful they'll keep both ARM and Intel Macs available for more than 2 to 3 years.

 

Even if people at Unity do know they're certainly not going to ruin Apple's party by breaking the news in advance.

 

Hmm... Is that a hint???

LOL

 

No. I'm just a customer. I wouldn't have access to that information to be able to drop hints.

All I'm saying is that even if Unity are privy to some of Apple's unannounced plans they wouldn't be able to tell us anything that might indicate what they are. They'd get in lots of hot water for that.

 

I see. Yeah, I wouldn't expect them to leak such info. Well, based on the rumors, I guess we'll know in less than 2 weeks time.

 

Apple usually treats software developers as crap, so if you can't go into their website and download tools for building for ARM Macs yet, then probably neither can Unity.

They'll obviously have some x86 emulation too. Apple likes to break compatibility, but releasing a Mac that cannot run any existing app on a short notice is too much even for them.

 

It won't be that short a notice. Apple will do something similar to when they switched to Intel. Announce it at WWDC and have a device for developers to develop and test their apps on. The OS is released the next year to the public.

 

No problem, Unity Editor will run on x86 only, osx builds will be treated as mobile. I love Apple's "good ideas".

(Obviously joking. A little. Or something.)

 

I'd love it if Mac ARM support came with or led into Linux ARM support too. The Raspberry Pi is slowly gaining more powerful tiers beyond just the initial $35 with the latest having 8 GB RAM which is much more in line with a desktop.

 

This has me pretty nervous. I love Macs just because of their OS and design and stuff, and I'm fully aware that half of the reason their OS is so smooth is that they don't concern themselves with long-term backwards compatibility like Windows does (Especially the transition form OS9 to OSX which was a completely different Operating System). However this puts me in a bind, as unless it's been secretly happening behind the scenes I can't see the Unity Editor being converted to ARM any time soon, which puts a limit on how long I can be using a Mac for development.

I could see them adding ARM build support easily (effectively an extension of the iOS exporter) but I wouldn't be surprised if we have to either use either Windows or an old Mac for the editor for a good few years. As much as Windows 10 has improved (still not quite up there with 7 but it's getting better all the time), and though I always keep a Windows laptop around for deploying to dev kits and such, I would be very sad to lose the option to develop on my Mac, it just makes my job a lot more enjoyable.

All that said, there is a somewhat plausible possibility that Unity have been working with Apple on this quietly under NDA for a while, like the "secret double life" of Mac OSX. When they revealed the iMac Pro they showed Unity running on it, and don't forget that Unity was originally a Mac program and didn't come to Windows until later. Plus Apple put a lot more pressure on developers to update stuff.

On the positive side, while it might take a few years, this could encourage a lot of developers to finally start ARMing up (so to speak!), which would not only make Macs into these crazy efficient machines but also give those ARM Windows machines a boost too. Laptops could become a lot faster and cooler and batteries could last much longer.

Patiently awaiting WWDC to hear more. I'll build a nice PC if I have to but I'd definitely rather get this rumoured new iMac, without feeling like my days in the Apple ecosystem are numbered.

 

Microsoft's transition from ME to XP was basically identical to Apple's transition from OS 9 to OS X. We saw a significant loss in backwards compatibility with the migration, and then we had a second one with the transition from XP to Vista.

Windows 10 is essentially Windows Vista but with years of updates. We haven't had a major revamp since then but we're going to need one soon because the internals are struggling to take advantage of the massive core counts (AMD is at 16 cores for consumer, 64 cores for HEDT). For comparison it's performance is around 80% that of Linux with the same app.

 

I'm gonna say not 'identical' but yeah, Windows XP switched home users over to Windows NT in the same way Mac OSX Switched Mac users over to NeXTSTEP. But a tonne of Windows 95/98 stuff still works on modern Windows without too much messing around. Try making a folder or file called CON or LPT1 on your PC, it won't let you, and that's left over from MS DOS even. It has a tonne of old stuff in there, there are even a couple of dialog boxes left over from Windows 3.1. XP did switch home users over to the Windows NT code base, but NT was developed alongside 9x and they were mostly compatible with each other. But yeah I get your point.

It's cool seeing the different attitudes towards it. A lot of what holds Windows back is old code, but it's also really helpful to have it in there because there's a lot of legacy software which still comes in handy in many situations, while Mac is much better optimised due to often dropping support for older hardware (and by extension, software) but it means we keep losing the ability to use certain things.

 

I don't imagine it would be too hard to build the Editor for ARM. I doubt there is any machine code in there. I imagine it would mostly depend on the tools/libraries used to develop/build the editor.

I wonder what the performance would be like. x86 processors (which have been more-or-less RISC processors at the core for like 2 decades now) are no where near as bad in terms of performance-per-watt vs ARM as some people think.

 

Well, they just announced that Mac will be transitioning to Apple Silicon, probably ARM. They support Intel apps through emulation though, which runs fast enough that it even supports AAA games, so Unity probably isn't going to have an issue on the new Macs.

They also announced a new type of Mac binary called Universal 2, which can run on both Intel and ARM Macs. Maybe a new build option is in order?

 

I’m seriously blown away. They have Maya and Shadow of the Tomb Raider seemingly running incredibly well on what I assume is the A12Z through emulation. A tablet chip! Seems like my initial concerns were unfounded. Doesn’t Tomb Raider run like garbage on even Intel’s latest integrated GPU? Still, hopefully Unity runs natively on day one or as soon as possible.

 

Yes.

Spoiler: Shadow of the Tomb Raider - Intel HD 630

That being said Intel's new Tiger Lake processors will have a completely revamped GPU. Below is a video showing Intel's upcoming Xe integrated graphics running Battlefield V at 1080p on high detail getting 30+ FPS.

Spoiler: Battlefield V - Intel Tiger Lake

 

Well, that was an amazing WWDC, and now we know.

Even though they showed AAA GAMES running perfectly through miracle x86 emulation (holy S***), I'm still a tiny bit wary of using an emulated program as a compiler. However after this I have enough faith that the people at Unity will be able to rebuild for ARM natively within too long, and until then I'm sure they'll be taking any potential emulation issues into consideration with their builds.

I don't need to update my Mac for a while yet, (and I'll be waiting for the new iMac form factor before I do) but I do feel much more confident now that when I do, it can be to another Mac. Incredible stuff. I wasn't planning to upgrade to one of the new Macs on day 1 anyway, but I felt that if I was going to have to move to PC, I wanted to start transitioning out of the parts of the Apple ecosystem I use ASAP and not continue to focus on Mac.

I would like to hear from Unity themselves to know they're planning a native ARM port, and until then building future x86 builds with this emulation in mind, hopefully working around issues if they arise.

Man, ARM CPUs in pro-level desktops, without the need of all that cooling and stuff... but with backwards compatibility with x86... the future of computers is going to be INSANE.

I have ordered a new, much better Windows laptop than my current cheap-ass one I keep on standby for dev kit deployment, since I expect Parallels and Bootcamp will be the one thing we lose, so while I'll be busting that out for Windows stuff in the future, I'm really thrilled to be able to continue using MacOS for my work.

 

https://www.tomshardware.com/news/apple-arm-intel-transition

 

Watch the WWDC Platforms State of the Union, they show Unity running on the new CPU!

Looks like Unity will support building Universal binaries as well. Check out this screenshot.

 

Is it native though?

 

Oh man, I missed it! I wish this thing would let you rewind to the beginning! I wanna see if this was native or what.

 

I don't think that is the important question, the important question is whether it can build native, which it shows it can. That's the big one. The runtime is native (and universal) so the editor is there, or will come eventually. Since it was shown in a screen in the keynote, undoubtedly there will be an announcement this week.

 

Or today. Hope you don't mind upgrading to Unity 2020.2.

https://blogs.unity3d.com/2020/06/22/unity-coming-to-apple-silicon/

 

Well, that’s reassuring. At least Unity should work day 1. Although Unity 2018 & 2019 is not going to work natively?

 

Heh.. we recently updated our projects to 2018.4.18f1. . Also I got a brand new top of the line MBP in March... so this whole ARM thing is likely about 2 years out for me. Which should ideally be enough time for the kinks to be worked out.

 

Yeah, this now has me very confident in continuing to use Mac. I do want to change is to go full SSD instead of Fusion Drive, but I'm waiting for the rumoured new form factor before I do that. I watched State of the Union and yeah, they show Unity working perfectly, fully native on ARM, hell it even appears to be enhanced by the metal API and faster in edit mode because of it all. That and Photoshop are the two main things I need so I'm all good but this x86 emulation is amazing too. I'm both blown away and relieved in equal amounts. It went from "I'm worried I might have to lose the pleasure of working on an iMac" to "Macs will now be better than ever for my work". Amazing!

 

That would be insane if Unity actually rebuilt their core engine to work with the new Apple Silicon. As much as I dislike what Apple did recently to app developers and the whole lawsuit drama, I'm still very excited for the future of Apple Silicon. Eventually IPads and Macbooks are going to be one and the same, so you'll literally be making next generation Unity games on a IPad. Microsoft has been pushing ARM on Windows 10 with Surface Pro X, which still has a lot of issues at the moment. But I definitely see a future where laptops will be in the same position as IPod Touches are today.

 

In case you missed it: https://blogs.unity3d.com/2020/06/22/unity-coming-to-apple-silicon/

 

I think it's a great fit

 

Quite possibly, yeah. We needed a change in CPUs for computers to move forward. Windows 10 for ARM has issues, for sure, but I'd say it's not Microsoft or Windows' fault, it's the lack of developer support. With Apple changing their entire lineup to ARM, developers will be forced to support it if they want to continue doing business, and my hope is that in the process they will also update to support ARM versions of Windows while they do it, and then everybody wins.

It's happened a lot with Apple in the past - In a way they 'take one for the team', for example choosing to nix flash support when making the iPad mean that web developers finally moved on from it and other laptops and mobile devices benefited from it too. They always get criticised for it but it always moves things forwards.

 

What does this statement even mean? In the past few years CPU performance for desktop computers has been moving forward at a more accelerated rate than it has in the previous 15 years. That's thanks to AMD, not Apple. Even though the instruction sets from older CPU designs are carried over, the internal workings of today's CPU's are nothing like the CPU's where the instruction sets first appeared, so there is "change in CPUs" already happening all the time.

I don't see how "everybody wins". Apple is making this move to get away from a fractured ecosystem of different CPU architectures, and to save paying Intel money for their products. Pushing Windows ARM accomplishes the opposite, by itself creating a fractured Windows environment of multiple CPU architectures where effectively none exists today. While Microsoft does make their own computers, they would be buying their CPU's from someone else regardless of what architecture they use, so Microsoft making such a push wouldn't give the company the benefit of in house development of the SoC that Apple will be seeing.

Generally when Apple moves things forward, it is when they are actually addressing an issue people care about. Flash was problematic from the user perspective, but had become this necessary evil in order to use the modern internet, but regular folks didn't like the performance issues, the near daily nagging about security updates, etc. So Apple took a stand on something customers really wanted a stand taken on.

But when I watched the keynote, I was just shaking my head. They kept repeating performance vs power consumption, performance per watt. But look at our space here, video games. Our customers literally do not care about performance per watt, at all. All they care about is overall performance, and performance per dollar (euro, etc). No gamer is deciding on hardware based on power efficiency. They are deciding based on frame rates and purchase price. If gamers primarily cared about performance per watt, they would all be running Intel Atom processors. So as far as in the desktop, ARM seems more like a solution looking for a problem rather than a solution for a problem. We will see I guess.

 

This is a very important point to make for anyone who hasn't studied microprocessor architecture design. Modern CPUs are built around the idea that complex instructions can be broken down into multiple simple instructions that are executed in the order that is most efficient by the processor. We call this microcode.

If you needed to compare it to anything to make it easier to understand it is essentially identical in principal to Rosetta.

https://en.wikipedia.org/wiki/Microcode

 

Moving on from requiring like 9 fans plus watercooling for a high-end rig will push computers forward. I stand by this.

When we can get gaming-grade laptops in a thin chassis with days of battery life without requiring a bunch of fans, everybody wins.

I can understand what you are saying, for sure. But think of it this way - The Nintendo Switch uses an ARM processor and has been able to run modern games in the form factor of a handheld, and it has been a huge success. I'd say for gamers it's less about 'performance per watt' than it is about 'performance vs heat/size'. Imagine what one of those CPUs could do in a cooled tower. The large form factor of a high-end gaming PC is based around the fact that x86 processors generate a lot of heat. ARM is built around using less power and generating much, much less heat. Until now, the focus has been on using ARM to miniaturise computers into spaces which are too small for cooling fans, such as smartphones, tablets, Nintendo Switch... but think about what could be achieved in something desktop-sized with cooling. I think it opens up a lot of opportunities and we will see a big change jump in computers soon. The PS5's design has been controversial because of how gigantic and unusual looking it has to be for efficient cooling. Similarly people have been weird about the Xbox Series X's vertical design. In both cases it was necessary due to cooling.

 

I hope you're exaggerating this on purpose because water-cooling is never a requirement. For that matter a large number of fans doesn't guarantee you have optimal airflow and cooling either. PCMR (PC Master Race) people love to do it to but it's much more of an aesthetic than an actual requirement.

Gaming laptops have minuscule battery life not because of fans but because of the actual graphics hardware that is necessary for a gaming experience.

Speaking of which the Nintendo Switch is a fantastic example against your arguments. For starters, it absolutely requires a fan to have adequate cooling and since it's a very small fan it has to have a high rotational speed but that has the downside of bringing much higher noise with it.

A PC will typically have fans between 120 and 140 mm spinning at around 400 to 600 RPM which in a room completely devoid of any other noise will be barely audible. My computer's fans at full throttle (~1,000 RPM) are barely noticeable despite the hardware being high-end (eg a GTX 1080 GPU). That said the fans have yet to hit full speed without being manually set to do so (I almost always check noise levels before concluding a build is finished).

Let's talk about battery life. Modern x86-based gaming laptops typically have a battery life of around an hour, and the Nintendo Switch while being ARM-based is barely an improvement with demanding games like Breath of the Wild only getting up to 2.5 hours. That's a far cry from a normal laptop's 7 to 9 hour battery life.

It's important to understand that ARM is not a magic bullet. Apple isn't switching because it's magically going to be everything they ever dreamed about. Apple is switching because it's the only way to have complete control over the hardware.

 

It's always been about performance-per-watt. The move from Power PC to Intel in 2007 was due to that, and so is the move from Intel to ARM in 2020. Longer battery life, less heat, what's not to like?

Plus, it's not like you're giving up on x86 apps just yet, there's a pretty good emulation layer.

 

I almost agree with this, except the reason why Apple ditched Flash. It wasn't because of the users, it was because another company gained control on the platform. They really didn't care about performance, they cared about kicking off Adobe. Otherwise their App Store wouldn't become The Thing so quickly. It is simple as that. They wanted control, so they gained control. Same thing is happening with this ARM-move. They are strongARMing (sorry, not sorry), because someone else is benefiting on their platform and now they have enough in-house know-how how to make a CPU what more or less workable.

Hahahahahaaaaaa.... that's cute. So you still believe in Apple that they can pull out performance from their a.... thin air.
Nice joke. Apple's "strong" laptops and hardware aren't gaming and aren't strong at all. They have ridiculously weaker hardware than the competition on every level. They can sell you like hotcake, but that doesn't mean it has performance, it just means they have one of the best marketing team on the planet.
And when something doesn't work on the platform they just ditch it and come out something jokingly bad replacement, so everyone shut up for a couple of years, because, well, you know, "it's new", they will iron out the quirks. (Metal, someone? No?)

edit(just for clarify what I implied but didn't write down): a lot of things are physically bounded. Like performance versus heat dissipation. You cannot really have a small chassis and big airflow at the same time unless you put some very small, very loud, very high-RPM vents. Or you can have a very weak hardware in there with very high temperature, like Apple does.
Gaming laptops aren't that big nowadays, just take a look at the latest MSI or Asus lineups. They made an excellent job imitating Apple's form-factor but keeping as much performance as possible. What Apple did is simpler, they don't give a flying ef about performance, they build a machine to look at and do average stuff without (barely) burning your skin.
Apple doesn't compete in the gaming or in the high-performance category.

 

Keep in mind we've been shown best case scenarios. There are many emulation solutions out there where the best case is fantastic and everything else is between mediocre and downright awful. Cemu, for example, has been labeled as a Wii U emulator but it was almost entirely built for Breath of the Wild and everything else was unplayable for a very long time.

 

You only need that kind of cooling with the max overclockers. There are no desktop CPU's which require more than an air cooled CPU fan and 1-2 chassis fans. What do you expect the cooling to look like with the same max overclocker folks get into ARM based desktop CPU's? Hint, it will look quite familiar.

We've been able to build laptops with that kind of battery life for decades. They don't currently exist, because the market demand doesn't exist. No matter the power consumption and battery technology, the industry has settled on approx 3-6 hours of battery life and sizes the batteries to hit that target. With improved power consumption, all you will see is the batteries shrink. If you want days of battery life, you'll need to start your own laptop company.

The Switch is a bad example. Portability puts it in a space where it really isn't competing with Xbox and Playstation. The Switch arguably competes more with gaming on phones than it does with consoles. Nintendo going all in with the handheld only Switch Lite makes this even more clear. Viewed in that context, the Nintendo Switch is actually the power hog in their space, the example of maximizing performance over efficiency. Exactly what I said gamers care about.

You're also making a leap here, from power efficiency to high end gaming performance. Apple's demo of Tomb Raider run at 1080p on low settings, as their best case example they could possibly put together for their keynote, tells me Apple isn't anywhere close to competing in high end gaming. Take pretty much any PC made in the past decade, throw in a low end 4 year old graphics card, and you'll get the same performance in Tomb Raider Apple just demoed. Where did Apple talk about being able to scale this up to compete at the high end of the performance scale with Intel/AMD? They didn't. If they thought they could, even far in the future, they would have said so.

 

This, and there are two major reasons for it. One, thin and light is the primary demand by the market and in pursuit of that something has to be eliminated. Batteries are one of the thickest and heaviest components in a computer and thus one of the first things to be penalized.

Two, airplanes have an imposed limitation on battery capacity due to their explosive nature. Modern electronics use lithium ion batteries. Currently the largest capacity you can carry onto the plane is 100 watt hours. Manufacturers build around this limitation which is why you never see a laptop with three digit watt hour batteries.

https://medium.com/solvaygroup/flying-with-lithium-batteries-7a2aa6ef7319

 

I didn't mean to imply Apple removed Flash for the purpose of helping their users. Only that is how after the fact the action has been viewed as moving things forward. Their decision to do so was almost entirely for business reasons. My understanding is it was done for 2 rather selfish reasons with the early iPhone releases. Firstly to have more control of their new platform, as you stated. Second was Flash was a memory hog, and Apple wanted to ship the iPhone stripped to the bone RAM wise to maximize profits by keeping costs low. The original iPhone ended up shipping with only 128MB of RAM, and since then pretty consistently shipped with approximately half the RAM of competing phones in the same class, and doesn't have page file support. The poorly optimized ads made with Flash that were all the rage at the time, didn't play well with the iPhone.

 

I honestly don't get all of the complaining here. What harm is there to switch to ARM? If the performance stays the same, what harm is there being more efficient? It's not like compatibility is a big issue too; even Unity has been cross-compiled already, and the x86 emulation is already shown to work with tools like Maya and even AAA games* with decent performance. It just sounds like complaining about change.

Like, some people here are saying more efficient power consumption will just lead to smaller batteries. Why? The industry has settled on 3-6 hours of battery life? This makes no sense. For high-end gaming laptops, yes (nothing is stopping them from increasing battery life as a selling point), but Macbooks have been pushing the 12-hour mark already.

Or using the Switch as an example that performance is more important than efficiency. Why not both? Imagine if the Switch had used an Intel chip instead. Now you've got the same performance, but less battery life.

*Yes, the game looked bad, but it looks more like a GPU issue. Which I agree it is and always has been a weak point for Macs.

 

Well I appreciate all the corrections about cooling and such details but I'm trying to say if a CPU runs cooler you can get more performance out of one at the same size. It ain't rocket science.

Pretty much, yeah. I was worried when I first heard about ARM (though as you can probably tell from my prior posts, I can see exactly why they want to do so, and not just because it'd be good for them), because I didn't want to lose support for the programs I need and have to switch to Windows for my main workflow, but WWDC and State of the Union, seeing Adobe and Unity already being ported and Rosetta 2 for everything else was a great relief.

I also think a lot of people just get really uncomfortable at the idea of x86 being phased out in some way or other because it'd be a big problem for PC gaming and the like. I'm not a PC gamer but if I was I'd hate the idea of my entire library suddenly not working on newer hardware. I doubt x86 will go away from PCs any time soon, one of the reasons to go with Windows is because it supports stuff going back a long way. But I'm just saying that if ARM-supported computers get the software people need, it'll be brilliant for the future of computers. They've been trying so long, with Windows 8 RT, Windows 10 S and now Windows 10 for ARM, the devices running those have always been so thin and light and cool, with battery life for days, but they've just not supported anything. A future where major software developers have to prioritise support for ARM CPUs is one where those devices can be worthwhile and if you ask me that is a win for everybody.

 

We haven't been shown anything beyond a demonstration and companies always show best case scenarios. We won't know whether this was a smart decision or a foolish one till we have access to the actual machines. Just like you can't know how a game will perform on a target device without the target device to test and profile against.

Clearly it is rocket science because your posts continue to show you have no idea how any of it works. There is only one way to increase efficiency without losing performance and that's by choosing a smaller process (eg 7nm -> 5nm). Every other method will require you to exchange one for the other (eg lower clock speeds for less heat and more efficiency).

Shrinking a process node comes with two choices. You can either choose to maintain efficiency while gaining performance or you can choose to maintain performance while gaining efficiency. Jumping from TSMC's 7nm process to their 5nm one can give you either 20% more efficiency or 15% more performance. You can't have both.

Apple almost always purchases the latest node from TSMC which is why they have been able to achieve the performance and efficiency that they have, but there is a limit to how far a process node will bring you. At some point it comes down to the way you designed your chip and how fast you're willing to run it.

https://www.anandtech.com/show/1272...-scaling-but-thin-power-and-performance-gains

 

Yes, exactly. After the lackluster efforts of Windows 10 S I feel like this is our actual real shot at switching over to ARM, thanks to Rosetta 2 and frankly just great execution on Apple's part. And I say this being a Windows user.

True. The first devkits are shipping out this week. However, it really feels like proven technology already; it's already in use in their other products, and the iPad is pretty darn fast for a device without active thermals. Given the benchmarks against Intel I can't imagine it'll be much different on Mac.

Probs out of my depth here but what about the difference in architecture?

https://stackoverflow.com/questions/14794460/how-does-the-arm-architecture-differ-from-x86

Quoted:

ARM is a RISC (Reduced Instruction Set Computing) architecture while x86 is a CISC (Complex Instruction Set Computing) one.

The core difference between those in this aspect is that ARM instructions operate only on registers with a few instructions for loading and saving data from / to memory while x86 can operate directly on memory as well. Up until v8 ARM was a native 32 bit architecture, favoring four byte operations over others.

So ARM is a simpler architecture, leading to small silicon area and lots of power save features while x86 becoming a power beast in terms of both power consumption and production.

About question on "Is the x86 Architecture specially designed to work with a keyboard while ARM expects to be mobile?". x86 isn't specially designed to work with a keyboard neither ARM for mobile. However again because of the core architectural choices actually x86 also has instructions to work directly with IO while ARM has not. However with specialized IO buses like USBs, need for such features are also disappearing.

End quote.

Also, while Intel does use RISC-like microcode, it still uses CISC instructions on the surface, adding complexity. Add in compatibility issues, and that is a whole lot of added complexity taking up space and power.

 

"If performance stays the same" is the whole problem. Your link compares the performance of the very best Apple has to offer against a quad core ultra low powered mobile processor from 6 generations ago, and surprise to no one, the best Apple has ever produced is apparently a bit faster. But Apple isn't introducing the new Mac into a market where the performance benchmark is set by Haswell based laptops. How does this compete with the 10980XE? Against the 3950X? Or against even the sub $1k budget build king, the Ryzen 3600? If it isn't on par, then what we have is a performance drop. Apple has so far released zero evidence performance will match current x86 offerings.

Because laptops have been here for decades, and that has always been the case. As efficiency of both electronics and battery storage increases, laptops get slimmer. You could cram the guts of a Macbook Air into a chassis the size of one of Apple's old Macbook Duo's, and have enough leftover room for probably a week's worth of battery. Why isn't Apple selling that product right now?

Please show me a Macbook review where it reports battery runtime is as advertised, and not several hours below the advertised spec.

Why not both? Because both is hard, and hard means expensive. If nVidia released a new low powered version of the 2080 TI, with identical performance but only 1/3 of the power usage, would you buy it over the regular 2080 TI if it were also double the price?

 

Yeah, effectively ARM has a bunch of 'shortcuts' to operations meaning it can do some things in a single operation that an x86 chip would take several to do, just cause ARM already has a bunch of stuff pre-baked. Think of it like asking two kids what 6x6 is, one of the kids goes "six plus six is twelve, twelve plus six is eighteen..." and so on until he gets to the answer, but the other kid has memorised his times tables and only has to do one calculation. Which of those kids is gonna get his math homework done fastest? I mean I'm dumbing that down and paraphrasing but you get the idea.

My friend, I appreciate your information but not your attitude.

 

For the record the actual term for these is "look-up table" (sometimes abbreviated to LUT). There is a major problem with this approach though and that is the method required to store them can often be slower than simply calculating everything from scratch. It has to be able to fit into a cache to be viable. Anything else is far too slow.

https://stackoverflow.com/questions/15452481/lookup-table-vs-runtime-computation-efficiency-c

Just as an example a hardware multiplier doesn't work in this fashion. A hardware multiplier has one stage for every bit in the numbers that it must multiply (eg a 16-bit multiplier will have a column of 16 rows of full adders). What you described with your example is much more akin to a software multiplier that has been long since discontinued due to inefficiency.

A look-up table for this example would be insanely inefficient. We would need to be discussing the calculations of sine, cosine, etc to make it worthwhile in any reasonable fashion, and then we're back to the problem of will it fit into a cache with all of the potential variations that need to be stored? For each of the difference functions?

If I'm reflecting an attitude it's simply because the posts I'm responding to read as if they were written with one. I tend to pick them up subconsciously as I post if I'm not actively trying to write in a neutral way.

 

I think you are missing the point, the big reason why Apple switching to ARM for their MAC products is such a big deal. It's the same reason why it's such a big deal that nintendo switch is getting all the big triple A games.

In the next couple years, the move to ARM will allow both Microsoft and Apple to create a robust desktop experience, that can be taken anywhere and used almost instantly. Obviously I'm not wishing to play next generation games on the new Apple Silicon. But having a portable robust desktop experience on a tablet is revolutionary. It's going to move computers forward for all users, regardless of whether they are using PowerPC, ARM, or X86/X64 Processors.