4 ms·

I'm not a Mac user, so I am not sure if there's something fundamental I am not appreciating, but I do not understand why releasing a high-specification desktop

by bhauer 10y ago

I'm not a Mac user, so I am not sure if there's something fundamental I am not appreciating, but I do not understand why releasing a high-specification desktop computer requires a year or more of design and planning. I do recognize that Apple has a reputation for making uniquely-designed cases, and I suppose there is some merit in that, but ultimately, a desktop computer gives a great deal of flexibility to design. On the easiest side of the spectrum, Apple could take the previous-generation desktop case (I believe Mac users lovingly call it the "cheese grater") and install a modern motherboard, CPU, memory, disks, and GPUs. There's virtually no design effort needed. Maybe a SuperMicro Xeon workstation board, a couple Xeon E5s, 64 or 128 GB of memory, a bunch of SSD and disk options, whatever GPUs work well with macOS. (These parts are roughly what I've used in my personal workstation with great success; but I don't run macOS.) Hackers do this every day that a Hackintosh is booted for its first time; but such hackers don't have the luxury of battle-testing device drivers for their grab-bag hardware. So they nervously worry about operating system updates. Obviously Apple would spend a month or two ensuring the drivers are bulletproof and boom, there's a new high-specification Mac Pro. If reusing an old design is simply too uncouth to contemplate, they could design a new nice, minimalist case sized for industry-standard components. Dozens of PC case designs would pass muster. I suspect most people who actually want a Mac Pro don't care if it's ultra compact or a mid-tower that sits on the floor. Why invest so much engineering effort in making it compact? Take away the size constraint and the required engineering effort essentially evaporates.

protomyth 10y ago
Not a defense, but I am guessing they are waiting for the next set of Xeon processors on Intel's roadmap. I understand Intel has announced 2018 Q1 for a group of processors.
- edvinbesic 10y ago
  Not only that, but I think that Apple is feeling the Pro user backlash and are really trying hard to say "sorry, but there is something awesome in the pipeline" where the pipeline is a few months old at most. It takes a while to get a product to market and this would seem to answer most of the criticism the pro market has had with the mac pro so far, the macbook pro is a different beast.
  - nihonde 10y ago
    Apple has weathered, and thrived during, much worse "backlashes" than the current one, which is itself lacking any evidence of being a real thing in terms of Apple's bottom line. AAPL is up over 50% since I bought it last year when everyone was claiming the iPhone was over.
- iwintermute 10y ago
  Also I imagine graphic card makers - it was a shame to see microsoft products with Nvidia 9xx series when 10xx is huge step up in performance/energy effectiveness. Plus something with storage - new intel optane would be nice for pro machine and so on.
tedmiston 10y ago
> I suspect most people who actually want a Mac Pro don't care if it's ultra compact or a mid-tower that sits on the floor. I think you're making the oversimplification they raise in the article about grouping all of the different pro users archetypes into the idea of "the pro user". (Some) developers might not care, but other pro types especially that make heavy use of peripherals certainly do.
- bhauer 10y ago
  > I think you're making the oversimplification they raise in the article about grouping all of the different pro users archetypes into the idea of "the pro user". (Some) developers might not care, but other pro types especially that make heavy use of peripherals certainly do. Be that as it may, the fact remains they could satisfy a contingent of the Mac Pro consumers in very short order if they delivered a traditional mid-tower case with modern-day commodity hardware components. Basically a Hackintosh that sits in an unremarkable but cleanly-designed case with internal components that Apple stands behind with tested drivers. A mid-tower Mac Pro is not some daring untested platform. It's what they had prior to 2013. If a segment wants a Mac Pro Mini, then maybe that is something to sink a year or more into. Flipped around, the unnecessary grouping they are doing is insisting that all Mac Pro consumers wait equally long for features that are (apparently) time-consuming to design and engineer. Yet we know from evidence directly in front of us that a segment of "pros" would immediately purchase a refreshed cheese-grater Mac Pro.
  - TheOtherHobbes 10y ago
    The distance between a working prototype and a shippable mass market product is huge. Of course you can go and hack together a high-spec Hackintosh in an afternoon. But that's the easy part. The hard part is setting up the parts sourcing at scale, organising the build channel, setting up the distribution/shipping channel, preparing the documentation and support, and designing the marketing message. You also need to deal with consumer regulations, safety testing, and EM compatibility in every territory you sell to. All of the above takes an impressive amount of time and money. If you're Apple, you also need to make sure the product fits with your future plans. And - probably - has some magic secret sauce you can tout to make it stand out more. Even a "simple" refreshed cheese grater is still a huge project. And I'm unconvinced a refreshed cheese grater would solve the real problem - which is now lack of trust in the brand, and lack of marketing momentum.
vizeroth 10y ago
Other than the appearance, a great deal of the design goes into cooling and noise reduction. Many Macs can do quite a bit without spinning up a fan, and the more powerful machines usually have more complicated mechanisms to make this possible. Yes, it's easier to do all of this with a bigger footprint, but Apple has a hard time pleasing people with the Mac Pro regardless of which direction they go, and at this point I don't think they would win over many fans by pushing out a standard tower.
- nottorp 10y ago
  No current Mac, with the possible exception of the trashcan, can do the "quite a bit" that a pro user needs without spinning up a fan. Keep any of them at 100% CPU for 5+ minutes and see for yourself.
hallman76 10y ago
It's an example of Apple's designers having more power/authority/influence over their engineers. Apple's earlier iphone & ipod cables are a more common example of this. They broke regularly because they were built to a design spec, not an engineering spec.
paulcole 10y ago
>but I do not understand why releasing a high-specification desktop computer requires a year or more of design and planning. You are VASTLY underestimating the planning/engineering/marketing/design/fabrication/logistics/etc. that a multi-billion dollar company undertakes to bring a new product to market, especially one whose brand is built (in large part) on design. Just rolling out something we've seen before isn't going to happen and even then it's not something that could happen quickly. It's really just that simple.
- scarface74 10y ago
  Apple throwing together a "Mac" in a bog standard PC case with Mac compatible hardware is something that Apple could do and would make most Pro users happy. Apple has in fact shipped "Macs" in standard PC cases and slapped OS X on them. They did it for the first developer x86 Mac machines. How long do you think it takes Dell to create a new case and slap hardware in it?
ScottBurson 10y ago
I am a Mac user, and I wholeheartedly agree: the cheese grater case is just fine. It doesn't take a standard motherboard, though. Still I can't imagine it would take them all that long to just update the boards they were using before.
kalleboo 10y ago
I understand that it can take a year to design and build a desktop computer (and the accompanying production line, parts procurement, etc), but what I don't understand is why it's taking a year from now. This is the company that famously had two teams building competing iPhone prototypes. One of the most valuable companies in the world, with a massive cash reserve and a skyrocketing R&D budget. Apple pundits are always talking about how Apple has prototypes of everything imaginable under the hood that they won't release if they don't think it's just right. And yet they were so blindsided on the Mac Pro that they didn't start developing a modular one until now?
rbanffy 10y ago
An expandable Mac is probably the least unique design Apple can do: its shape is dictated by the PCIe slots and the connections on the back. Airflow and noise management are issues, since Apple has little control over the fans people put on expansion boards, but, again, it's a compromise: you wanted a "pro" machine, it'll not be quiet. Having said that, it'd be fun to see a Cray-like thing that wraps the PCIe boards around a flex-PCB bus going around a central core (CPUs, memory, heat management) while keeping the mass storage between the empty wedges created that way.
simonbarker87 10y ago
Even with the billions of dollars Apple has in the bank the absolute quickest they could turn around what you are describing would be 9 months at a guess. ATP guess around 200,000 Mac Pros are sold a year - the first year of a new version after so long could easily be 1 million units. Getting that production and logistics line set up is a big undertaking even if using old cases etc. For the extra 6-9 months that a new design would take it's totally worth it for them as will hopefully give them a modern design that is good for 5-10 years that meets the needs of the 2017+ pro market, using a chassis that was designed for the 2008+ needs would be a step back for Apple and pros in general. Apple like to push the envelope and while they certainly do make mistakes (I count the trash can as a mistake but not the Touch Bar) there's no point in them going ultra conservative.
simonh 10y ago
Apple sells their computers at a premium and to do so they need to provide premium features and quality. Lets say it takes 6 months to design and test a new system at peak design efficiency using current parts. If you want a cutting edge system at launch though you need to design and test it to take advantage of the CPUs, GPUs, chipsets, connector technologies etc that will be cutting edge then, not now. That’s a lot more complex and takes much longer. Also the launch schedule is not just driven by your own product development, it also has to be synchronised with the component manufacturer’s product development. That can easily push out the schedule even further because you’re not working with current known, well understood parts. You’re working with pre-production specs and early prototype samples. Most manufacturers don’t get privileged acces to these, but you can bet your bottom dollar Apple does. Apple also have a history of collaborating with parts manufacturers to specify custom variants of parts or even designing entirely new parts, such as the custom timing controller chip in the original 5K iMac. Working at that level takes even more time.

by bhauer 10y ago

protomyth 10y ago

Not a defense, but I am guessing they are waiting for the next set of Xeon processors on Intel's roadmap. I understand Intel has announced 2018 Q1 for a group of processors.

edvinbesic 10y ago
Not only that, but I think that Apple is feeling the Pro user backlash and are really trying hard to say "sorry, but there is something awesome in the pipeline" where the pipeline is a few months old at most. It takes a while to get a product to market and this would seem to answer most of the criticism the pro market has had with the mac pro so far, the macbook pro is a different beast.
- nihonde 10y ago
  Apple has weathered, and thrived during, much worse "backlashes" than the current one, which is itself lacking any evidence of being a real thing in terms of Apple's bottom line. AAPL is up over 50% since I bought it last year when everyone was claiming the iPhone was over.
iwintermute 10y ago
Also I imagine graphic card makers - it was a shame to see microsoft products with Nvidia 9xx series when 10xx is huge step up in performance/energy effectiveness. Plus something with storage - new intel optane would be nice for pro machine and so on.

tedmiston 10y ago

> I suspect most people who actually want a Mac Pro don't care if it's ultra compact or a mid-tower that sits on the floor. I think you're making the oversimplification they raise in the article about grouping all of the different pro users archetypes into the idea of "the pro user". (Some) developers might not care, but other pro types especially that make heavy use of peripherals certainly do.

bhauer 10y ago
> I think you're making the oversimplification they raise in the article about grouping all of the different pro users archetypes into the idea of "the pro user". (Some) developers might not care, but other pro types especially that make heavy use of peripherals certainly do. Be that as it may, the fact remains they could satisfy a contingent of the Mac Pro consumers in very short order if they delivered a traditional mid-tower case with modern-day commodity hardware components. Basically a Hackintosh that sits in an unremarkable but cleanly-designed case with internal components that Apple stands behind with tested drivers. A mid-tower Mac Pro is not some daring untested platform. It's what they had prior to 2013. If a segment wants a Mac Pro Mini, then maybe that is something to sink a year or more into. Flipped around, the unnecessary grouping they are doing is insisting that all Mac Pro consumers wait equally long for features that are (apparently) time-consuming to design and engineer. Yet we know from evidence directly in front of us that a segment of "pros" would immediately purchase a refreshed cheese-grater Mac Pro.
- TheOtherHobbes 10y ago
  The distance between a working prototype and a shippable mass market product is huge. Of course you can go and hack together a high-spec Hackintosh in an afternoon. But that's the easy part. The hard part is setting up the parts sourcing at scale, organising the build channel, setting up the distribution/shipping channel, preparing the documentation and support, and designing the marketing message. You also need to deal with consumer regulations, safety testing, and EM compatibility in every territory you sell to. All of the above takes an impressive amount of time and money. If you're Apple, you also need to make sure the product fits with your future plans. And - probably - has some magic secret sauce you can tout to make it stand out more. Even a "simple" refreshed cheese grater is still a huge project. And I'm unconvinced a refreshed cheese grater would solve the real problem - which is now lack of trust in the brand, and lack of marketing momentum.

vizeroth 10y ago

Other than the appearance, a great deal of the design goes into cooling and noise reduction. Many Macs can do quite a bit without spinning up a fan, and the more powerful machines usually have more complicated mechanisms to make this possible. Yes, it's easier to do all of this with a bigger footprint, but Apple has a hard time pleasing people with the Mac Pro regardless of which direction they go, and at this point I don't think they would win over many fans by pushing out a standard tower.

nottorp 10y ago
No current Mac, with the possible exception of the trashcan, can do the "quite a bit" that a pro user needs without spinning up a fan. Keep any of them at 100% CPU for 5+ minutes and see for yourself.

hallman76 10y ago

It's an example of Apple's designers having more power/authority/influence over their engineers. Apple's earlier iphone & ipod cables are a more common example of this. They broke regularly because they were built to a design spec, not an engineering spec.

paulcole 10y ago

>but I do not understand why releasing a high-specification desktop computer requires a year or more of design and planning. You are VASTLY underestimating the planning/engineering/marketing/design/fabrication/logistics/etc. that a multi-billion dollar company undertakes to bring a new product to market, especially one whose brand is built (in large part) on design. Just rolling out something we've seen before isn't going to happen and even then it's not something that could happen quickly. It's really just that simple.

scarface74 10y ago
Apple throwing together a "Mac" in a bog standard PC case with Mac compatible hardware is something that Apple could do and would make most Pro users happy. Apple has in fact shipped "Macs" in standard PC cases and slapped OS X on them. They did it for the first developer x86 Mac machines. How long do you think it takes Dell to create a new case and slap hardware in it?

ScottBurson 10y ago

I am a Mac user, and I wholeheartedly agree: the cheese grater case is just fine. It doesn't take a standard motherboard, though. Still I can't imagine it would take them all that long to just update the boards they were using before.

kalleboo 10y ago

I understand that it can take a year to design and build a desktop computer (and the accompanying production line, parts procurement, etc), but what I don't understand is why it's taking a year from now. This is the company that famously had two teams building competing iPhone prototypes. One of the most valuable companies in the world, with a massive cash reserve and a skyrocketing R&D budget. Apple pundits are always talking about how Apple has prototypes of everything imaginable under the hood that they won't release if they don't think it's just right. And yet they were so blindsided on the Mac Pro that they didn't start developing a modular one until now?

rbanffy 10y ago

An expandable Mac is probably the least unique design Apple can do: its shape is dictated by the PCIe slots and the connections on the back. Airflow and noise management are issues, since Apple has little control over the fans people put on expansion boards, but, again, it's a compromise: you wanted a "pro" machine, it'll not be quiet. Having said that, it'd be fun to see a Cray-like thing that wraps the PCIe boards around a flex-PCB bus going around a central core (CPUs, memory, heat management) while keeping the mass storage between the empty wedges created that way.

simonbarker87 10y ago

Even with the billions of dollars Apple has in the bank the absolute quickest they could turn around what you are describing would be 9 months at a guess. ATP guess around 200,000 Mac Pros are sold a year - the first year of a new version after so long could easily be 1 million units. Getting that production and logistics line set up is a big undertaking even if using old cases etc. For the extra 6-9 months that a new design would take it's totally worth it for them as will hopefully give them a modern design that is good for 5-10 years that meets the needs of the 2017+ pro market, using a chassis that was designed for the 2008+ needs would be a step back for Apple and pros in general. Apple like to push the envelope and while they certainly do make mistakes (I count the trash can as a mistake but not the Touch Bar) there's no point in them going ultra conservative.

simonh 10y ago

Apple sells their computers at a premium and to do so they need to provide premium features and quality. Lets say it takes 6 months to design and test a new system at peak design efficiency using current parts. If you want a cutting edge system at launch though you need to design and test it to take advantage of the CPUs, GPUs, chipsets, connector technologies etc that will be cutting edge then, not now. That’s a lot more complex and takes much longer. Also the launch schedule is not just driven by your own product development, it also has to be synchronised with the component manufacturer’s product development. That can easily push out the schedule even further because you’re not working with current known, well understood parts. You’re working with pre-production specs and early prototype samples. Most manufacturers don’t get privileged acces to these, but you can bet your bottom dollar Apple does. Apple also have a history of collaborating with parts manufacturers to specify custom variants of parts or even designing entirely new parts, such as the custom timing controller chip in the original 5K iMac. Working at that level takes even more time.