You're Apple. You just built a 40-core SoC for Mac Pro. Now what?

[dmccloud]

As far as I know, the M1's memory bus do not use PCIe at all. That would be disastrous for bandwidth performance. I do not think any CPU's memory busses go thru PCIe. CPUs have direct access to their DRAMs via dedicated memory bus lanes.

This is correct. What Apple widened was not the memory lanes as deconstruct60 has claimed, but the decode pipeline. Since ARM instructions are a fixed length, it's much simpler to widen the decoders than it is on x86 (where instructions are variable length). AMD has flat out admitted that a 4-wide decoder is about as wide as x86 can go due to the complexities of variable-length instructions. Since the RAM is on the same die as the CPU, it is a direct access model instead of via an external bus.

 

[jdb8167]

This is correct. What Apple widened was not the memory lanes as deconstruct60 has claimed, but the decode pipeline. Since ARM instructions are a fixed length, it's much simpler to widen the decoders than it is on x86 (where instructions are variable length). AMD has flat out admitted that a 4-wide decoder is about as wide as x86 can go due to the complexities of variable-length instructions. Since the RAM is on the same die as the CPU, it is a direct access model instead of via an external bus.

The RAM on the M1 is on the same chip carrier but not on the same die. The RAM chips are separate components from the SoC.

 

[senttoschool]

Lion share of Microsoft's revenue comes from office and corporate/business users (over 70%). They want you to use MS Office and Active Directory, and offering Windows in the cloud is a smart way of making sure you stay in their revenue-generating ecosystem no matter which device you prefer. In addition, office/business work is well suited to remote computing, since display updates are rare and the latency of data transmission over internet can be tolerated.

Appel however wants you to use their physical devices. They care less about you running their OS itself, they want you to be part of the experience they set up by matching hardware and software features, as that is where their profits come from. Services like iCloud, the new Xcode cloud make perfect sense, since they add value to the seamless "Apple magic" and allow you to use your device more comfortably. But the experience is still focused on you and the device in front of you. I am not sure how macOS in the cloud fits in here as it won't be able to deliver that low-latency user experience Apple wants to be known for. Take video editing for example. Current M1 Macs are excellent video editing machines, because their memory architecture allows them to say super responsive while working with high-res content that brings nominally faster machines to their knees. You just won't get the same experience from the cloud computing: bandwidth and latency are simply not where you want them to be.

I can imagine some sort of "compute server" in the cloud, where you app runs locally on your Mac but will offload the data to the Apple server to have it processed there, but again I am not sure how useful that would be. Data transmission would probably take longer than the time you'd save from using a more powerful cloud machine, so that leaves you with only a few very niche workloads.

Anyway, that's just my thoughts. I don't claim to be any kind of authority on these matters.

Seems like you're more "Team Federighi" and I'm more "Team Stauffer".

This email chain was written in 2017 by the way when cloud gaming was still a myth and Geforce Now was still 3 years from launching. Their stance could have changed entirely within the last 4 years now that they've seen more of what other companies are doing via the cloud and they know more about their Mac chips.

I think there is a middle ground in which Apple integrates local and cloud environments seamlessly. If you control the OS and the Cloud, you can do things like selectively use local hardware for latency-sensitive things and the cloud for acceleration. You can hit a button in the OS and have a hybrid environment where some applications are streamed from a 256-core Apple Silicon SoC but your Notes app is still running locally. You can even use one desktop as local and the other as Cloud and all you need to do to switch between them is a 3 finger swipe.

Eventually, I think cloud centers will be built everywhere (edge) to reduce latency and more and more things can be done in the cloud. If gaming can be streamed from the cloud, most things can.

Sticking to pure local compute is a losing game long term though. You have to think that Apple is smart enough to see the trends.

 

[leman]

Seems like you're more "Team Federighi" and I'm more "Team Stauffer".

Well, Craig is a very smart guy

I think there is a middle ground in which Apple integrates local and cloud environments seamlessly. If you control the OS and the Cloud, you can do things like selectively use local hardware for latency-sensitive things and the cloud for acceleration. You can hit a button in the OS and have a hybrid environment where some applications are streamed from a 256-core Apple Silicon SoC but your Notes app is still running locally. You can even use one desktop as local and the other as Cloud and all you need to do to switch between them is a 3 finger swipe.

I admit that there is some attractiveness to this, but actually getting there is going to be tough. And of course, high-end Apple Silicon availability is one of the core problems. I doubt that Apple will be able to produce enough of high-end chips (not to mention that they will be very expensive) to offer a cloud computing solution any time soon.

Sticking to pure local compute is a losing game long term though. You have to think that Apple is smart enough to see the trends.

That I agree with. Currently, Apples plan seems to involve offering a much better value proposition than anyone else, but at some point they will have to think bigger.

 

[senttoschool]

Well, Craig is a very smart guy

He is indeed but this was 4 years ago. The cloud has been growing so much and in so many different ways. The cloud advantages are too great to ignore: sharing hardware so that it gets full utilization to cut cost, access anywhere, managed by someone else, and upgrades without big upfront capital investments.

Latency issues can be minimized as more data centers are built nearer to the edge.

I admit that there is some attractiveness to this, but actually getting there is going to be tough. And of course, high-end Apple Silicon availability is one of the core problems. I doubt that Apple will be able to produce enough of high-end chips (not to mention that they will be very expensive) to offer a cloud computing solution any time soon.

I think Apple will finish the transition, then surprise everyone with some kind of MacCloudOS in 2023 or 2024. By then, supply and demand should be settled.

They can start with renting out the biggest SoCs first for businesses then slowly trickle down to consumers who need extra power once in a while.

That I agree with. Currently, Apples plan seems to involve offering a much better value proposition than anyone else, but at some point they will have to think bigger.

They will have to for sure. This isn't the Steve Ballmer Microsoft. This Microsoft is formidable. And Google & Amazon are aggressive. Then you have the Chinese internet giants as well. Apple can't be too stubborn.

 

[Krevnik]

If gaming can be streamed from the cloud, most things can.

I think the irony here is that gaming is both an ideal starting point, and a terrible one, for streaming applications themselves over the internet.

On the one hand, there’s a large market you can chase that doesn’t have a high end gaming rig, but you can deliver 4K to their TV. On the other, the more real-time nature of gaming, and the fact that some games are built around almost twitch-level responses to what happens on screen means latency and video compression artifacts are the hurdle that has to be overcome. And instead of focusing on trying to understand the target market and delivering games suited for the target market and would provide a good experience while streamed, you are seeing FPS games get pushed a lot which is super strange. I’m not surprised it’s been rough.

I think one reason I’d be more in favor of “Team Craig” here, though, is more that while the cloud is great for a lot of things and provides a lot of efficiencies, it tends to undo some of the democratizing effect of providing access to computing hardware to the masses by centralizing data and computing into the hands of organizations that set Terms of Service that you must abide by, which are inherently more strict than your average copy of Windows or macOS, and ties you to subscriptions. While I’m sure the convenience outweighs the costs for many, I don’t think I’ll ever be ready for a future where my devices are reduced to dumb terminals.

 

[ChrisA]

The Mac Pro is a very niche market relative to iPhones, iPads, and low-end Macs. Investing a significant chunk of your Apple Silicon R&D into the smallest market doesn't seem to make sense.

My guess is hat Apple will drop the entire idea of pro level hardware and software. Far more people watch movies than make them.

Just like with Aperture. Apple will claim to be 100% committed to it until one day when they aren't. There will be no warning, just one day, a simple announcement. Don't laugh, they did it before.

 

[Jorbanead]

My guess is hat Apple will drop the entire idea of pro level hardware and software. Far more people watch movies than make them.

Just like with Aperture. Apple will claim to be 100% committed to it until one day when they aren't. There will be no warning, just one day, a simple announcement. Don't laugh, they did it before.

This doesn’t make sense to me. Apple had their chance to ditch the pro market back in 2013 when the Mac Pro bombed and many pro’s slowly left Apple. Instead, Apple created a pro workflows team, doubled down on a new design, realized said design was taking longer than anticipated, reworked the iMac to support Xeon processors, released numerous major updates to final cut and Logic, went against their MO to release a modular Mac, created custom hardware like their PFGA Afterburner card, worked with AMD to build custom GPU enclosures. Why go through all this effort and spend all this money on RnD just to ditch the entire pro market a few years later?

Deprecating aperture is not at all the same as dropping the entire pro market.

 

[leman]

My guess is hat Apple will drop the entire idea of pro level hardware and software. Far more people watch movies than make them.

Just like with Aperture. Apple will claim to be 100% committed to it until one day when they aren't. There will be no warning, just one day, a simple announcement. Don't laugh, they did it before.

Apple is not stupid. They know that abandoning the pro segment — no matter how small it is — will mean slow death for their entire business. Having a powerful workstation is not as much about profits as it is about reputation. In many regards, Mac Pro is a symbolic product for Apple, and a very important one. It's what makes people take Macs seriously, even if they would never buy a Mac Pro themselves.

 

[Kung gu]

There is zero "first mover ' advantage here for Apple. The M-series being native boot only macOS makes it even less so. ( Linux support dependent upon the jailbreak hack to get going isn't something that critical 24/7/365 business critical workloads are generally going to bet the farm on. It is an amusing hack that works for some. It isn't primary time though. If Apple closed the jailbreak vector in 3rd or 4th generation update that is a dead end. )

There is no need for jailbreak to boot other kernel's on M1 Macs.

Does Apple allow this? Don’t you need a jailbreak?​

Apple allows booting unsigned/custom kernels on Apple Silicon Macs without a jailbreak! This isn’t a hack or an omission, but an actual feature that Apple built into these devices. That means that, unlike iOS devices, Apple does not intend to lock down what OS you can use on Macs (though they probably won’t help with the development).

From here: https://asahilinux.org/about/

 

[jlc1978]

Apple is not stupid. They know that abandoning the pro segment — no matter how small it is — will mean slow death for their entire business. Having a powerful workstation is not as much about profits as it is about reputation. In many regards, Mac Pro is a symbolic product for Apple, and a very important one. It's what makes people take Macs seriously, even if they would never buy a Mac Pro themselves.

I'm not so sure. I'd guess most Apple user's consider MacBook Pro the "Pro" version and are unaware of what the Mac Pro is used for in various industries. If it was that important for their reputation you'd see it highlighted a lot more.

Apple may dabble in the high end market and have a flagship product but not having one would not make much, if any, difference in their success. As long as it is a profitable niche they'll make one but if it starts losing money they'll dump it.

 

[senttoschool]

I'm not so sure. I'd guess most Apple user's consider MacBook Pro the "Pro" version and are unaware of what the Mac Pro is used for in various industries. If it was that important for their reputation you'd see it highlighted a lot more.

Apple may dabble in the high end market and have a flagship product but not having one would not make much, if any, difference in their success. As long as it is a profitable niche they'll make one but if it starts losing money they'll dump it.

I think/hope Apple has a much bigger plan for the Mac. Yes, Apple Silicon will increase consumer/semi-pro market share. But I believe that Apple wants a big piece of the enterprise pie. That's where a lot of the big money is. In order to get a big piece of the enterprise pie, Apple needs to offer solutions from entry-level desk workers (MBA/Mini) all the way to Mac Pros. If you're going to be a "Mac shop", you need to cover use cases from top to bottom.

Our company is a Mac shop. All engineers get 16" MBPs. And we have Mac Pros to help us build our apps faster.

 

[leman]

I'm not so sure. I'd guess most Apple user's consider MacBook Pro the "Pro" version and are unaware of what the Mac Pro is used for in various industries. If it was that important for their reputation you'd see it highlighted a lot more.

Apple may dabble in the high end market and have a flagship product but not having one would not make much, if any, difference in their success. As long as it is a profitable niche they'll make one but if it starts losing money they'll dump it.

I think that these dynamics can end up being incredibly complex. True, Mac Pro is at best a niche product, but it’s users carry weight that reverberates through the rest of the ecosystem. A lot of professional software that Mac is known for for example relies on this small group.

It’s like with gaming hardware - the number of people who own a 3090 RTX is virtually zero, and yet it’s these flagship GPUs that create momentum and draw attention.

I think/hope Apple has a much bigger plan for the Mac. Yes, Apple Silicon will increase consumer/semi-pro market share. But I believe that Apple wants a big piece of the enterprise pie. That's where a lot of the big money is. In order to get a big piece of the enterprise pie, Apple needs to offer solutions from entry-level desk workers (MBA/Mini) all the way to Mac Pros. If you're going to be a "Mac shop", you need to cover use cases from top to bottom.

I very much doubt that Apple is going for the enterprise market. That is firmly Microsoft’s domain and frankly, I think getting Macs to be accepted as gaming machines is going to be a much easier job than grabbing the enterprise. Apple has probably realized it a while ago, seeing how they’ve all but abandoned the server platform. But Apple is definitely targeting workstation market, a d they have great potential there with Apple Silicon.

 

[Jorbanead]

But Apple is definitely targeting workstation market, a d they have great potential there with Apple Silicon.

Agreed 100%

I'm not so sure. I'd guess most Apple user's consider MacBook Pro the "Pro" version and are unaware of what the Mac Pro is used for in various industries. If it was that important for their reputation you'd see it highlighted a lot more.

Sure many consumers may not know of the Mac Pro. But plenty of actual professionals are fully aware.

Having worked in the audio industry, and working in video, UX, photography and other creative fields, macs are everywhere in these professional markets. Sure not everyone owns a Mac Pro, but everyone I interact with is aware of them. Mac has had tons of success in audio/visual professional markets for decades, they have invested heavily in developers over the last few years, and their workstations seem to be catered towards this. That’s why it’s not crazy to me to think Apple could start investing in gaming as another market to tap into. That would certainly open up a whole new level of Mac users. I know a lot of people think that even suggesting such a thing is ridiculous, but It would take a company like Apple to disrupt the Nvidia/Microsoft duopoly in computer gaming and it fits much nicer with their already established “creative professionals” and “developers” markets.

 

[GrumpyCoder]

I think/hope Apple has a much bigger plan for the Mac. Yes, Apple Silicon will increase consumer/semi-pro market share. But I believe that Apple wants a big piece of the enterprise pie. That's where a lot of the big money is. In order to get a big piece of the enterprise pie, Apple needs to offer solutions from entry-level desk workers (MBA/Mini) all the way to Mac Pros. If you're going to be a "Mac shop", you need to cover use cases from top to bottom.

Our company is a Mac shop. All engineers get 16" MBPs. And we have Mac Pros to help us build our apps faster.

That is never going to happen. No matter what hardware they offer, they'd have to offer proper support for hardware as well, worldwide, which they can't handle. Look at what Dell and Lenovo are doing, that's the level they need. That also means a much higher level of custom systems. It is simply not acceptable to have a full day of downtime on any system. Same day service, no matter what happens, I expect any faulty system to be repaired and back up and running within hours.

That's the level of service Apple would need. The closest you'll get is via a large reseller which is stocking replacement parts for everything and is taking the risk. Good luck with that.

 

[Joe The Dragon]

I think/hope Apple has a much bigger plan for the Mac. Yes, Apple Silicon will increase consumer/semi-pro market share. But I believe that Apple wants a big piece of the enterprise pie. That's where a lot of the big money is. In order to get a big piece of the enterprise pie, Apple needs to offer solutions from entry-level desk workers (MBA/Mini) all the way to Mac Pros. If you're going to be a "Mac shop", you need to cover use cases from top to bottom.

Our company is a Mac shop. All engineers get 16" MBPs. And we have Mac Pros to help us build our apps faster.

for enterprise they some kind of server them.
Maybe just an wait to run mac os in an VM On any base hardware.

Apple Silicon mini server with IPMI / hot swap storage and dual power.

 

[senttoschool]

I very much doubt that Apple is going for the enterprise market. That is firmly Microsoft’s domain and frankly, I think getting Macs to be accepted as gaming machines is going to be a much easier job than grabbing the enterprise. Apple has probably realized it a while ago, seeing how they’ve all but abandoned the server platform. But Apple is definitely targeting workstation market, a d they have great potential there with Apple Silicon.

That is never going to happen. No matter what hardware they offer, they'd have to offer proper support for hardware as well, worldwide, which they can't handle. Look at what Dell and Lenovo are doing, that's the level they need. That also means a much higher level of custom systems. It is simply not acceptable to have a full day of downtime on any system. Same day service, no matter what happens, I expect any faulty system to be repaired and back up and running within hours.

Business - Mac

Work on all your projects at once with apps that run faster. Finish that email you started on your iPhone, without charging all day. Mac does that.

www.apple.com

Enterprise - Business

Apple hardware, software, and services work together to supercharge business connection, collaboration, and creativity.

www.apple.com

Apple wants the enterprise market. It's exactly the type of market a $2.5 trillion-dollar company needs to get into.

For the longest time, Apple couldn't compete with the likes of Dell/Lenovo in value since they offered the same Intel/AMD/Nvidia hardware but at a discount. Apple Silicon changes that equation just like it does for consumer markets. Apple has an opportunity to add more value than Dell/Lenovo.

Support infrastructure can be built. Many enterprises use third-party support companies like CDW without going directly to Dell/Lenovo/HP anyways.

 

[leman]

Business - Mac

Work on all your projects at once with apps that run faster. Finish that email you started on your iPhone, without charging all day. Mac does that.

www.apple.com

Enterprise - Business

Apple hardware, software, and services work together to supercharge business connection, collaboration, and creativity.

www.apple.com

Apple wants the enterprise market. It's exactly the type of market a $2.5 trillion-dollar company needs to get into.

For the longest time, Apple couldn't compete with the likes of Dell/Lenovo in value since they offered the same Intel/AMD/Nvidia hardware but at a discount. Apple Silicon changes that equation just like it does for consumer markets. Apple has an opportunity to add more value than Dell/Lenovo.

Support infrastructure can be built. Many enterprises use third-party support companies like CDW without going directly to Dell/Lenovo/HP anyways.

Ah, I think we just use the terminology differently. Of course, Apple wants businesses to buy Macs. But „enterprise“ market is something different: that’s user and group management, centralized IT management, communication, team productivity, data sharing etc. Apple has basic tools for managing Macs, but that’s about it. They used to offer enterprise productivity features, but they were always half baked and recently Apple completely dropped all the efforts on that front instead simply focusing on personal computers.

 

[Apple Knowledge Navigator]

Apple is not stupid. They know that abandoning the pro segment — no matter how small it is — will mean slow death for their entire business. Having a powerful workstation is not as much about profits as it is about reputation. In many regards, Mac Pro is a symbolic product for Apple, and a very important one. It's what makes people take Macs seriously, even if they would never buy a Mac Pro themselves.

Agreed - this is likely why they placed so much emphasis on the form/presentation of their current MP/XDR etc. It's this 'over engineering' that makes Apple somewhat stand apart.

 

[GrumpyCoder]

Apple wants the enterprise market. It's exactly the type of market a $2.5 trillion-dollar company needs to get into.

For the longest time, Apple couldn't compete with the likes of Dell/Lenovo in value since they offered the same Intel/AMD/Nvidia hardware but at a discount. Apple Silicon changes that equation just like it does for consumer markets. Apple has an opportunity to add more value than Dell/Lenovo.

Support infrastructure can be built. Many enterprises use third-party support companies like CDW without going directly to Dell/Lenovo/HP anyways.

What you link to is (small) business, not enterprise. Apple always tried to bring hardware into small business. Enterprise is different. Say you buy 7- to 8-figure in hardware every year, you need this to be running 24/7 and with guaranteed 24h or same day support on-site. Where is that feature from Apple? Where are the servers? Where the redundant power supplies? Where is my 2 PB of redundant storage? Not even talking about custom software options. Apple wants to sell exactly the same hardware to businesses as the end-user consumer market. That's not going to work.

And if you think Apple Silicon is going to change that, you have not paid attention to the enterprise/business market in the past few decades. Where is all the specialized ERP software running on AS? If anything people won't change their software or update if not for security reasons. That's why they cling onto old versions of Windows, they don't want to make changes. For well over a decade, I've developed custom software solutions on Apple hardware for the healthcare market and consulted in other fields. If anything, the "switch" to Apple used to be much easier back then, than it is is today.

 

[senttoschool]

Apple Considered Launching Cloud Gaming Service

Apple has discussed launching a cloud-based gaming service to rival the likes of Xbox Cloud Gaming and Google Stadia, according to Bloomberg's Mark Gurman. In his latest "Power On" newsletter, Gurman explained that Apple has considered launching a cloud-based game streaming service like its...

forums.macrumors.com

Hint: Apple isn't going to use Xeon to power their cloud gaming service.

 

[Joe The Dragon]

Apple Considered Launching Cloud Gaming Service

Apple has discussed launching a cloud-based gaming service to rival the likes of Xbox Cloud Gaming and Google Stadia, according to Bloomberg's Mark Gurman. In his latest "Power On" newsletter, Gurman explained that Apple has considered launching a cloud-based game streaming service like its...

forums.macrumors.com

Hint: Apple isn't going to use Xeon to power their cloud gaming service.

well apple arm is not cloud enterprise even rackmount ready.
Needs lot's of ram
Duel power
Hot swap raid 1 or better disks
IPMI
high speed networking / pci-e slots

 

[zakarhino]

Bumping this thread as it's now been confirmed by Moore's Law Is Dead that Apple are indeed working on server chips for in-house backend infrastructure with the possibility of selling these chips to third parties.

Congrats @senttoschool your theory looks like it might be true.

 

[Jorbanead]

Bumping this thread as it's now been confirmed by Moore's Law Is Dead that Apple are indeed working on server chips for in-house backend infrastructure with the possibility of selling these chips to third parties.

Congrats @senttoschool your theory looks like it might be true.

Do you have a link to the source? I can’t find this “confirmation”

 

[Kpjoslee]

Bumping this thread as it's now been confirmed by Moore's Law Is Dead that Apple are indeed working on server chips for in-house backend infrastructure with the possibility of selling these chips to third parties.

Congrats @senttoschool your theory looks like it might be true.

Confirmation by Moore's Law Is Dead is not a confirmation at all.