Andy Hertzfeld on the Demise of the Mac Pro | Hacker News Reader

Andy Hertzfeld on the Demise of the Mac Pro

plus.google.com

82 pointswyclif79 comments

Supposedly Tim Cook has said that updates for "pro customers" are coming next year.

http://www.macrumors.com/2012/06/11/david-pogue-new-imacs-an...

"Our pro customers are really important to us...don't worry as we're working on something really great for later next year." - Tim Cook

That's what Apple was promising for years before the FCP update and look how that turned out.

They don't have to do nearly as much work to satisfy the Mac Pro market. It's mainly just keeping the hardware up-to-date so the enthusiasts and high end users aren't using 2 or 3 year old parts.

I thought most of the FCX grouses had been responded to.

Some of it. The last update did quite a bit ( see http://support.apple.com/kb/HT4589 ), but they still have a ways to go to replace the previous version.

Some points, yes. Unfortunately, at least for my they have stated that they won't be bringing back the features that kill it for me (tape support.)

Tape's probably still used for a lot of work and will for some time, but it will be like bringing back the floppy disk:

http://collider.com/film-camera-production-ended-arri-panavi...

http://arstechnica.com/gadgets/2012/06/the-silver-screen-no-...

>That's what Apple was promising for years before the FCP update and look how that turned out.

Yes, how did that turn out?

It turned out as a great program, that Apple spend millions and years to rewrite from scratch to have a clean future proof codebase.

If you want to bring an example of Apple not caring about the pro market, Final Cut Pro X is NOT it. If anything, it's a counter-example.

Only idiots that don't understand software engineering complained. Ie, that you have to cut some corners, since you cannot have everything from the old version + new features, in a 1.0 release and ship in any reasonable timeframe.

So, yes, it didn't have ALL the features from the previous version. But soon after it was introduced, it got 2-3 updates, with major missing features (like multi-camera editing) added. And it's just in version 1.0 (post rewrite).

Compare to what happened with Netscape. Navigator wasn't going nowhere. But Mozilla took a lot of years and didn't bring everything Navigator had from day one. But thanks to that rewrite --which took like 5 years to produce something stable and usable--, we still, in 2012, have a stable, fast, Firefox. If that rewrite had not happen, there would be no Firefox today, Navigator would have died unable to compete with IE (much less with Chrome and Safari).

I honestly don't think this is enough. My video editing pals are dumping FCP in droves and there is no heat behind FCPX (updated or not) at all. And where there's no heat, there's no shop talk. Where there's no shop talk, there's no business needs discussion, and where there's no business needs discussion there's no sale.

Apple lost control of the dialogue here.

I don't see that. Actually professionals (in general, but also people I know) don't update that often. Think 5-6 years timescale. They need a stable base, and they don't even upgrade OSes or computers. So, in all the time FCPX is out, they are still 1 or 2 FCP versions behind. They might get on the FCPX bandwagon on v2 or even 3.

In another field I know even better because my brother works on it, professional music studios still use Logic 7 and 8, a lot even use G5 (G5's man!).

All the hoopla about FCPX is from amateur and semi-pro guys, that can jump around from program to program without care. It's not as if they have (much) work to do. See also the second version of the FCPX impressions (after the first updates), from some pros in Philip Bloom's site:

http://philipbloom.net/2012/02/07/fcpxeditors/

Considering the power of the new macbook pro, im guessing the update will be a convergence of the iMac and Mac Pro lines into a single desktop beast machine

I doubt it. Pro buyers want to select a display and it would be silly to build expandability into an iMac form factor.

I read that Intel is no longer manufacturing the CPUs the Mac Pro used, so Apple had to do an unscheduled upgrade, which would explain this horrible spec-bump.

You guys are totally missing the point. Although I'm typing this on a new Chromebook that does an amazing job at serving me the web I still need to run Pro Tools, which really does require a lot of hardware once projects become complex enough. Although a huge percentage of what we currently do with computers can effectively be accomplished on top of rather modest hardware, what is left is of critical importance: video/audio production, computationally intensive research, etc.

It can be argued that one could just do those things on Windows / Linux instead, but that is a pretty poor solution for anyone who is already used to Mac OS X or who owns other Apple hardware such as their Thunderbolt display.

I think the support status of OS X hardware (and software) is reason enough to switch to Windows or Linux for anything computationally intensive. They seem to throw it away at a whim.

Dells Precision machines are cheap, last a LONG time and always bang up to date. Never will you worry about being left behind.

Windows 7 is supported until 2020! Linux is supported until whenever you want it to be (you have the source).

Being used to something doesn't really cut it if you ask me.

>It can be argued that one could just do those things on Windows / Linux instead

Unless you are stuck in a Mac Only workflow. I've been using Logic since it was called Notator on an Atari ST. It did used to be available on windows, but I never used it there. That has long since been killed, and not it is Mac-only. There are plenty of other people in similar situations. I could go out and learn Cubase or similar, but I'd rather not.

Apparently someone has had a reply to an email sent to Tim Cook promising a bigger upgrade to the Mac Pros next year.

http://www.macrumors.com/2012/06/11/david-pogue-new-imacs-an...

I posted this in the other thread, and I've been thinking about it more overnight.

Yes, this was a pretty "lame" update.

At this point, there are two things Apple could be doing:

1. The Mac Pro is EOL. Dead. Gone.

2. The Mac Pro is alive and well, and they're working on a new model.

Now, if the truth is option #1, why would they release this "lame" update now? There is no reason to. When they wanted to kill the xServe, they just killed it. Done.

I think the more likely scenario is option #2, but for whatever reason, they have not been able to build an Ivy Bridge/USB 3/Thunderbolt beast to their liking yet. At D10, Tim Cook clearly said they don't build things to "arbitrary schedules, or price points... they just build the best products they can". So a few months ago when it became clear the "new" Mac Pro was not going to be ready any time soon, they green-lighted this "lame" bump, to hold over until the new beast is actually ready.

I think this "lame" update shows a heartbeat for the Mac Pro, as a hold-over until something newer is actually ready to go prime-time.

(The difficulty I always think of is will the Thunderbolt port(s) be on the gfx card, or the motherboard? Hmmm)

I bet they're tired of the shape and are trying to figure out how to do something different.

Ideally they'd come up with something that could somehow work better in either rack mount or desktop usage.

Software keeps getting more nimble. Storage is moving to the cloud. There's really nothing super surprising about this move. No one had heard of thunderbolt a year ago and now all of the sudden it's a necessity?

The large majority of consumers will probably be impressed by the next gen Mac. That is the direction the entire line will be heading in and Apple clearly wants to make the differentiation clear, instead of marginally different and making people indifferent between the NGMBP and MBP.

These are not just creative decisions, they're marketing decisions, and so far Schiller hasn't done much wrong.

It's notable because creative professionals were once a very important market for Macs, and yet now Apple isn't even pretending to care about them.

And creative professionals always could use more power. Even a current top of the line PC isn't enough.

Updating a Mac Pro from Westmere to Sandy Bridge E is absolutely trivial compared to creating the next MacBook Air. You can even do it at home with a Hackintosh - they are almost entirely commodity PCs in the same old boxes they've been in for years.

And yet it hasn't been done.

Wouldn't creative professionals, at least those working in design and video, be pleased with the Retina display? Where else are they going to find a display of this caliber to work with?

Yes, but it's on a 15" screen. I haven't had a 15" screen on my desktop since 1994. Resolution is important, but for many tasks, so is size. The new MBP has a 2880x1800 screen, the old Apple 20" display had a 2560x1600 screen. So, while the MBP has a few more pixels, it's not that many more pixels.

Having seen neither, I can's speak to quality issues on the two screens.

I hope this doesn't mean Apple are moving away from people in the creative industries. They'v gotten good at making stuff for grandmothers, five year olds, yuppies and programmers etc. etc. That's really impressive and they did a great job.

But I think their roots in the creative industries were always one of the contributing factors to the quality of their stuff. The idea that a mac is for doing things like video editing and music production is IMO a proxy for making it good at other things too.

Who is to say that the domain of people in creative industries are restricted to users of Mac Pros.

Creativity is not bound by hardware.

Nobody. Creative people use all sorts of things.

As a product though the Mac pro targets (not exclusively or exhaustingly) a section of creative professionals: designers, music industry people, video editors. The design choices & margins produce a price that basically targets a 'need what they need regardless of price' segment. They've been doing so for a long time. In the Apple dark ages these people were a big chunk of mac users. You can still see the vestiges of this heritage in modern macs, eg iLife: Consumer versions of what their professional (in that space) users use.

It is if your version of creativity requires rendering large amounts of media. You could render it on a Macbook Pro in 5 days, or on a Mac Pro in a few hours (actual time may vary).

Powerful machines, servers, and rendering farms exist for a reason. A Macbook Pro is not powerful on the same scale of what the Mac Pro should be.

I'm a creative professional. I do most of my work on an iPad 2 during the day. I use Adobe Proto and Photoshop on it, OmniGraffle, Coda Lite, and Apple Pages and Numbers. When I get home, I get my files from Dropbox/Adobe.com/iWork.com, and use my 2004 PowerBook G4 to turn it into an InDesign document, if it hasn't been published on the web already.

I also have a recent Mac mini, but even though it's the beefiest machine that I own, I only use it to view movies.

Seems like there are a few different, not-necessarily overlapping use cases for Mac Pros.

There's the people who need lots of CPU. There's the people who need lots of RAM. There's people who want lots of fast disk. There's people who need to use specialized cards.

Trying to serve all at the same time dictates the shape of the machine. (For example, reduce space for cards, and you can make a 1U rack mount.)

Thunderbolt opens up the possibility of replacing the Mac Pro with something akin to a beefed up iMac or maybe a stretched shoebox-shape Mac Mini, with high-end CPUs, lots of RAM slots, and basic graphics and disk, with external Thunderbolt chassis available for disks or expansion cards.

If the main unit + disk chassis + card chassis cost the same as a similar Mac Pro, this would be a significant savings for those customers who only need the main unit and zero or one external chassis.

Unfortunately this is like a high end Commodore64: it starts simple but soon you have a rats nest of cables separate power for all the external devices. I want less cables not more. Also, I don't think thunderbolt cables can be screwed in, so you run the risk that a cable could get tugged and pulled out.

I'm thinking like an 8-drive chassis, and maybe 3- and 6- slot card cages, rather than lots of single-purpose external devices.

Rats nestery need not be too bad. Thunderbolt from the Mac to the drive chassis, then thunderbolt daisy-chained to the PCI card cage chassis. Power cords for the drive chassis and the PCI card cage chassis, if they can't be powered by thunderbolt. (Naturally, how ratty the PCI card cage gets would mostly be determined by what cards it contains and what are plugged into it.)

Not sure what the cable length limits are, but it might be possible to spread these out a fair distance, especially once optical-based Thunderbolt happens. Being able to put the PCI cards close to what they're connecting to, with just a single Thunderbolt cable running back to the computer, could be a cable management win, versus snaking multiple cables to a computer.

I'm sure someone could come up with a screw-in variant of Thunderbolt, but even so, I don't think Mac Pros have any screw-in connectors as it is, apart from the video cables screwing into the video card(s).

Thunderbolt PCIe expansion boxes are already in the works, such as this $979, 3-slot one: http://www.magma.com/thunderbolt.asp

My issue with the lackluster update is that I purchase a computer to last me ~5 years. The last mac pro I purchased was the octocore mac pro. I use it to run Logic, Reason and associated plugins, and find I can comfortably run 24-48 track projects quite easily so I havent really felt a pinch to upgrade. I have 16GB ram and 4 TB of disk space in it, so I think it still exceeds what most desktop/laptops are capable of to this day. Back to the point, since I want my purchase to last, it better have the current state of the art on the day of purchase so that accessories and upgrades are easily added.

I think the current Mac Pros are probably for customers like my last job, a medical school neuroscience lab that used Mac Pros to run the experiments at the five experiment rigs. If one of those Mac Pros needs to be replaced, something roughly equivalent would suffice - there's no pressing need for thunderbolt or USB 3, the goal is simply to replace the hardware ASAP and have a functional rig for doing the work.

>I use it to run Logic, Reason and associated plugins, and find I can comfortably run 24-48 track projects quite easily so I havent really felt a pinch to upgrade.

24-48 track projects? With slight "freeze" of some heavy reverbs and such, you could probably run those on a Dual Core MacBook Pro.

Absolutely, I have tons of headroom for more. However, just because I can means I should :) Amplitube usually is the biggest CPU drain I use, and I dont really freeze tracks that often.

If you want a high-end machine, buy a PC.

I had to buy a laptop recently and chose a PC because I could get 32G of RAM in a PC laptop. For my work, I need that much.

I want a high-end machine, and I want a Mac. Please stop telling people how to spend, or how they should want to spend their money.

The Mac Pro is still a powerful machine. We just didn't get what we hoped for.

Hackintosh is your answer.

Build the computer of your dreams, drop a bit of cash on a nice case if that matters to you, and install OS X.

It's really easy these days, especially if you have similar hardware to legit Macs.

Apple's the one that has strong ideas about how you should spend your money. PC makers listen and deliver a diversity of products.

Now, I'll grant that Apple products have advantages in terms of stability because there is less finger pointing when things go wrong -- you could get a mac laptop that sleeps correctly about 3-4 years before windows laptops got it right.

I don't see any real U.I. advantage on the Mac; if you try to put your photos in a movie, for instance, the Mac will drive you absolutely up the wall because it wants to make you look like a Shmuck who can't turn off the Ken Burns effect.

What you did end up choosing? I would love to get a laptop with that much memory.

I'm going to guess the Lenovo W520/W530 or a Dell Precision Mobile Workstation. Those are the major laptops I know of that come with 4 DIMM slots (32GB now, 16GB last year because 8GB DIMMs were crazy-expensive).

Currently sat at my w520 and it's a brilliant piece of kit. I've always been a thinkpad fan - although I can appreciate the draw of the MBP.

It's little touches like the spill-proof keyboard, carbon fibre reinforced plastics, trackpoint, and amazing thermal management (even with an i7 in there the bottom never gets hot, only mildly warm).

HP EliteBook Mobile Workstation can also use 32GB of RAM. We have a few of them at work and we're happy with them.

I suspect that in most cases where people need a very high end machine for work, price is not as much of a factor as running the software (including OS) that you want.

Having very demanding needs in hardware specs but flexible on operating system is probably a relatively rare occurrence.

This is an honest question, not a criticism, to the people who are disappointed with the Mac Pro offering: is a two-processor, 12 core machine at 3 GHz with 64 GB of RAM and 4 storage bays (up to 1 TB HDD, up to 512 GB SSD per bay) not enough workstation for you? If not, what do you use your workstation for, and what are your needs?

I ask because that seems like a hefty machine to me. The max specs on the Mac Pro are almost as much as the development nodes in the cluster I have access to, and those nodes host about 10 developers per node.

One of the biggest shortcomings is the inability to get a modern top-of-the-line graphics card in an official supported Mac Pro configuration. I know people with heavily GPU-dependent workflows that have switched to Windows workstations for this purpose specifically.

I think there's two factors here. First, the cost. If you're only going to offer 2-year-old hardware, it's hard to justify keeping the same price (especially when the base price was a bit high to begin with). Second, the Mac Pro is lacking hardware that would allow it to be future-proof (SATA III, USB3, Thunderbolt, new socket, etc). Basically, you're buying a dead-end of a computer at this point.

Thunderbolt is the #1 thing. Apple's own $1000 display is not supported by its high end desktop computer, which is just absurd.

They still sell their other $1000 display, the display port version, which is supported by the Mac Pros that they are selling.

True, but Thunderbolt is the newer technology. I recognize there might be some bias going on here, but the truth is that buying a display port display right now feels a bit like buying a S-VGA CRT to me.

I just recently bought the Display Port model. I just couldn't justify the additional expense of upgrading my MacBook to a Thunderbolt-supporting model.

It wouldn't be as bad, if it could easily be daisy-chained off a Thunderbolt monitor.

Unfortunately it appears that you need another Thunderbolt device after a Thunderbolt monitor, before you can connect a display port monitor.

(ie, Thunderbolt Display -> Thunderbolt Hub -> DisplayPort Cinema Display)

I think Hertzfeld mentioned the lack of thunderbolt and USB 3.0. So it's probably not the power that's lacking.

It's a reasonable question. My answer is: Thunderbolt.

The only Apple display that can simulate pixel-for-pixel an iPad Retina display is the (discontinued) 30", and only in landscape orientation. The only way to test a Retina Mac app today (short of buying a new Macbook Pro) is to view it double-size.

I know bigger-than-15" Retina displays would be super-expensive today, but I assume the eventual Retina external displays will be (like the current 27" display) Thunderbolt-only, and that Thunderbolt requires CPU/chipset support (right?), which makes it sound to me as though no current Mac Pro will ever show (correctly-sized) Retina content.

That's not a dealbreaker for me today, but Mac Pro buyers tend to buy them to last for several years. Nobody really wants to pay more for a new expandable machine that only supports an end-of-life expansion port. A Thunderbolt Mac Pro would at least stand a good chance of being Retina-upgradeable (with a new graphics card).

As someone who hasn't been following the news in these last few days, I got very depressed after reading this headline - which of course (silly me) turned out to be complete and utter sensationalistic bullshit.

I agree the silent almost-not upgrade is not a good sign, but it may simply be due to the release schedule of Intel's server chip. Anyway, the Mac Pro still exists and it's still a good machine. By these standards I have an ancient 2008 Mac Pro and it's absolutely good enough for anything I can throw at it.

It was true that the Mac Pro didn't receive any updates in the prior two years because there simply wasn't a new Intel chipset to update to...

But Intel's Sandy Bridge Xeon chipset and new E5 processors have now been put for a few months, with Dell's Precisions and other workstations now using them.

Maybe there is a bigger update in the works and Sandy Bridge wasn't worth the investment, so Apple is holding out for the next step. I don't know Intel's roadmap. But I do know Apple has no problems throwing out products it deems superfluous, so I've been worried about the Mac Pro's future for quite a while. As a multi-screen kind of guy, I will definitely leave the platform once the only upgrade path left is the iMac.

The headline wrongly suggests that the Mac Pro was canceled. It wasn't.

"As a multi-screen kind of guy, I will definitely leave the platform once the only upgrade path left is the iMac."

You can connect two external screens to an iMac, making a total of three screens.

http://gigaom.com/apple/new-27-inch-imac-supports-dual-exter...

I want my display and my computer separate. I want to connect non-Apple displays. Sometimes I want more than three screens. I want to be able to upgrade my graphics card, which I have done on my Pro 2 times. I'd love to be able to chain an arbitrary number of Thunderbolt displays together trading refresh rate for real estate but that's just not how it works, in fact it will probably never work this way.

Yes. Headless Macs are important for those who need high end displays. Depending on your needs, Apple really never was in that market.

Apple's displays are now all glossy, have dropped from 16:10 to 16:9, and most have such color inconsistency across the panel that they're impossible to calibrate. An iMac without a yellow band somewhere on the screen is like a unicorn these days.

"Yes. Headless Macs are important for those who need high end displays. Depending on your needs, Apple really never was in that market."

They were, but it's been a while. Quadra 840AV (1993) , PowerMac 9600 (1997).

At the time, if you wanted multiple screens, the Apple Macintosh platform provided flexibility that you couldn't get on the Windows NT side. Competitors were Silicon Graphics, Sun, and NeXT, but their workstations were way more expensive.

It's obvious that Apple has moved more towards consumers, and financially, it has worked for them. But I too would like Apple to give some love to professional desktop users.

http://en.wikipedia.org/wiki/Macintosh_Quadra_840AV

http://en.wikipedia.org/wiki/Power_Macintosh_9500

Buy a 3-slot Thunderbolt PCIe expansion chassis, fill it with top-end Mac-compatible video cards, and connect it to a Mac Mini via Thunderbolt.

I think you can daisy-chain six of these Thunderbolt-PCIe boxes, which would give you, what 36 monitors?

And not thunderbolt monitors, they'd be whatever the gfx cards you use support, so probably DVI or HDMI or DisplayPort.

It's a good idea but the Mini isn't really in the same category as far as processing power and memory capacity go. And this extension chassis is ridiculously expensive for no reason. Still, it's a good alternative if/when the Pro goes out.

Yes, as I note elsewhere, what might be nice would be a "MacMini Pro", with appropriate CPUs, RAM slots, reasonable default graphics, and a couple hard drive/SSD slots, in a sleek, compact box with adequate power supply and cooling.

I'm sure Apple could produce extension chassis at a lower price point than 3rd parties, if they wanted to, but I doubt they want to get into that business.

I'd mention the new high end hdpi display has 3 additional outputs. Two thunderbolt / displayport, one hdmi, for a total of four.

When people suggest Apple has an evil plan to impose the walled garden on our laptops and desktops, and turn all of our computers into appliances, I roll my eyes. But their neglect of the Mac Pro lends those arguments at least a little bit of credence.

On the other hand, despite Apple's huge cash reserves, I'm sure their human capital resources are limited. With iPhones, iPads, and MacBooks as wildly successful as they are (and for good reason, IMO), it would be really, really stupid not to allocate most of their resources to these products, and push for the major updates there.

I think we did not see new Mac Pros and new iMacs because Apple wants to update them to use a retina display as well. I think it's more likely that Apple expects the displays to be available soon than that Apple doesn't care about these lines anymore. If Apple wants to update the big screens, it will need to update the iMac and the Thunderbolt displays at the same time. Since the MacPro should be able to use the Thunderbolt retina display, it probably should be updated at the same time. I might be wrong that it's the screens they are waiting for, maybe it's the graphics cards, but I am sure all those product lines will be updated at the same time and they all will get Retina and probably within the next 6 months.

Looks like the Mac Pro is alive indeed.

http://www.macworld.com/article/1167247/cook_apple_planning_...

Quoting email from Tim Cook:

Franz,

Thanks for your email. Our Pro customers like you are really important to us. Although we didn’t have a chance to talk about a new Mac Pro at today’s event, don’t worry as we’re working on something really great for later next year. We also updated the current model today.

We’ve been continuing to update Final Cut Pro X with revolutionary pro features like industry leading multi-cam support and we just updated Aperture with incredible new image adjustment features.

We also announced a MacBook Pro with a Retina Display that is a great solution for many pros.

Tim

I wonder what Apple uses for development internally - I suppose they have clusters for fast compilation. Sucks if you are an external developer who needs to build large codebases, if this Mac Pro is the best you can buy.

At this point, I wish Intel had a non-Xeon multi-socket capable chip. The Xeon is serving two masters (enterprise server and pro workstation) and does a really poor job of it.

Its a shame Apple will not release a non-Xeon Mac Pro.

I think you mean "Mac Pro".

I fixed it. See how pervasive Apple marketing is? ;-)

Jesus christ, this sense of entitlement is fucking nuts.

Do you have a computer? Can it access the web? Is it less than 5 years old? Congratulations, you have a pretty amazing computer that should be able to do most things pretty reasonably.

I mean Jesus, some people make it seem like it's a personal attack when corporations don't pander to all.

What does this have to do with "can it access the web?". We're talking about top-of-the-line, most expensive, consumer destop PCs designed to do heavy tasks like video processing or scientific calculations.

Some people actually use computers to get real work done and not simply for browsing the web. Doing 'most things reasonably well' doesn't even come close to being good enough in a lot of areas.

"Testify Brother Testify"

the rules of tecnical computing

1 there is no such thing as too much RAM

2 there is no such thing as too much IPC

3 there is no such thing as too much DASD/Disk/Backing Store

4 there is such a thing as to little budget :-(

4 there is no such thing as a screen too big (or too many screens)

Andy Hertzfeld's opinion is relevant because he was one of the key people behind the original Macintosh released in 1984.

For some background: http://folklore.org/ProjectView.py?project=Macintosh&cha...

Can you render a 3 hour Final Cut Pro movie in a reasonable amount of time?

That's the Mac Pro use case.

Reply on news.ycombinator.com