A Mac Pro Mini | Hacker News Reader

A Mac Pro Mini

marco.org

76 pointsmh_96 comments

Apple has abandoned any users who need real power. They have abandoned enterprise. They have abandoned video, 3D, DAW. My company, which would love to maintain our Mac fileservers and OpenDirectory, but we are being forced, kicking and screaming into the open arms of Windows Server, Active Directory, and ExtremeZ-IP. We don't need "desktop-lite". We need server-grade multi-core workhorses to run CPU-intensive workflow automation services (graphic arts, print production, high-end color stuff), but can't get them.

And their iMacs and Mac Mini's? Yeah, we use them as servers because we have no choice. We actually need Thunderbolt now. But they are buggy as hell. The built-in network interfaces on them drop for no apparent reason, and need to be re-started. Three separate boxes do this: 2 iMacs and one Mini. We have actually resorted to putting SmallTree PCIe gigabit ethernet cards (Intel Pro chipsets) into Thunderbolt PCI express chassis, in order to get a stable ethernet stack. Those never drop on us. How can Apple get something as basic as a stable Ethernet port so wrong? No load testing. No desire to fix their drivers or firmware, whichever is responsible.

While I realize its not 'apple approved' it is very easy and cheap to build an extremely powerful 'hackintosh' as long as you pick the correct hardware. See http://www.tonymacx86.com/

I am glad that is possible (and tempted to do it at home every time I go to plug in a thunderbolt device into my Mac Pro, and then have to go oh yeah, god dammit...) it is not just not 'apple approved' it is not legal.

In most normal business environments where obeying the law is a required practice and bootlegging software is not allowed, building a hackintosh isn't a solution.

Not to mention, supporting any more than a few Hackintosh machines would be a nightmare for any serious organisation.

There is no need to bootleg any software and no it is not 'Illegal'. You can and should buy a copy of OSX which you can do on the app store without much issue. It is however, against the 'terms of service' which is not a legal issue but one of support / danger of being 'cut off' by an update.

That said, keep in mind that much of the work done to allow this is done by people disenfranchised by Apple's lack of a decent, upgradable, 'pro' model. In other words your often being supported by people in the same position as you.

I'm sure if we did that then our auditors and compliance officers would love the fact that we chose a system with full knowledge that we'd be breaking the contract we bought it under.

It is not a criminal issue, but it is certainly a legal issue. Even if you buy the install media at the Apple store, you don't legally have the right to install the copyrighted work on your hackintosh.

I'm not saying there are likely to be real-world consequences, just that most normal businesses have policies in place to make sure they comply with legal and contractual obligations, which make hackintoshes a non-starter.

I should not have said you have to "bootleg" the OS, though; as you point out, that is wrong. You can simply buy it and then install it in violation of the license terms -- which is not the same thing at all.

Even if it were allowed, it would not be supported by the vendor in terms of warranty support. when you're a 27x7 org, you don't need excuses from vendors as to why your setup is not a supported configuration.

LDAP issues. Apple will not support it. Driver issues, the HW mfg will not support the HW when weird things pop up.

"when you're a 27x7 org,"

I assume you mean 24x7 :P

I can imagine a not-too-distant future where we have cloud hosting available[1] from flocks of solar-powered UAVs[2] loitering around the International Date Line, thus finally paving the way for >100% uptime.

[1] http://www.abc.net.au/science/articles/2005/06/10/1389516.ht...

[2] https://en.wikipedia.org/wiki/NASA_Pathfinder#Helios_Prototy...

In general relativity and closed-timelike-curve-time-travel there is a concept of a universal clock, against which >100% uptime is impossible

Wasn't one of the points of the OP that Apple is doing a poor job of supporting the systems they do sell?

I honestly don't understand how this issue is any different than the numerous companies whom purchase generic hardware to run linux on, there isn't much support from either end, yet they manage to create some impressive systems from this 'unsupported' hardware/software.

How is running OSX on 'unsupported hardware' that much different than running a postgres server on EC2? Would the same company force their devs into using MySQL because RDS exists? If your reliant on computers beyond 'MS Office', and you don't have a dev team to deal with it, your going to lose.

Granted there will be those companies who would never 'take this risk' but, frankly, that is an issue of company culture, not one of suppliers.

There are a lot of companies out there doing amazing AV work using mac's as front end machines despite the lack of a mac pro update, how are all these folks pulling it off without Apple's hardware support?

I know of a 20-person animation shop which uses hackintoshes exclusively.

I really wish that Apple would go back to licensing its OS for clones: say they charged $2-300 minimum or 10% on a $1500 minimum selling price. This would enable almost reasonably priced powerful hackintoshii, while not cannibalizing their current products too badly.

I can vouch for the terrible ethernet controllers on the new iMacs. Random dropped connections, reboots required after waking from sleep.

Interestingly, we haven't had any problems with 21.5 inch models, only the 27's (although we've got more 27's).

Another Apple-in-the-enterprise sysadmin here. I confirm the same issues as mentioned above.

Why not Linux or the business distros from Red Hat and Novell?

Because Adobe's Creative Suite does not run on Linux.

We abandoned Linux for Mac, because Samba was horrendously slow when running on a large case-sensitive filesystem; because netatalk (we MUST have AFP at our company, given our industry) is horrendously slow an large filesystems of any type due to the CNID database it maintains, and it doesn't support spotlight searching from Finder clients (evidently there is now alpha support - too little too late); and because Linux does not integrate with a large network of Macs nearly so well as a proper OpenDirectory installation.

AND because we are running Mac-based workflow automation software for which there is no equivalent in Linux. For example, it drives Adobe Creative Suite in an automated fashion, as well as other print-industry specific tools.

> Apple has abandoned any users who need real power.

I still think Apple might bring at least 1 more generation of Mac Pro's on the market. Main reason would be for development of both Mac OS X as well as iOS. I think developing operating systems is best done on powerful hardware, mainly due to long compilation times.

But perhaps all this is wishful thinking. Apple could use Xgrid to build their operating systems on a bunch of Mac mini's of course.

I just nabbed an old iMac G4 with 10.4. It's clearly designed for a more power user; it openly advocates having your own website running on it.

Whereas my 10.8 MBP tries to play at being a tablet.

I'm basically done with Apple after the 10.7/10.8 direction of OSX.

I agree with your sentiment, everything is still OK right now. But if I am forced to use Apple's equivalent of a "Metro" interface, I am done and going to Linux.

Let me sing your hosannas. I'm in this exact same boat, watching a fleet of of (MacPro1,1)s pop their videocards while the 'test' iMac 27" struggles to keep up with work life despite being utterly tricked out.

There is simply no acceptable substitute for a Mac Pro, and we're going to bail if Apple won't spend the effort.

Well I dont think that a retina iMac will suffice. I had a mac pro for my home music studio. The point of it was to be a beefy machine that was plenty expandable had lots of disk space and was quiet for the horsepower it provided. Mac pros fit that bill perfectly. it was easy to load them up with fast server class hard disks plug in your audio interface, any DSP boards, dongles or whatever else was required and then you forget about it (that is, until you see your electric bill the next month...)

I got the 2008 octocore Mac Pro and did just that and ran both mac os and windows on it until I finally sold it last year(regrettably to help pay for personal circumstances as well as a lack of time for music production) and never had a single issue with it.

THAT is what I spend the extra money on, I want a beefy computer that will last me 5 years. That is what the PRO audio crowd wants them for. I would imagine its not much different in the graphics/video arena either.

A Xeon class CPU is quite a bit different than the mobile/desktop class CPU Apple puts in their other systems.

I think Macro is being a bit dramatic in that a stagnant Mac Pro means the end of an era and I call bullocks!

Intel's Xeon chips do not advance at the same pace and have themselves stood still. Look at what Dell and HP are selling, they're basically the same CPUs (albeit you can buy one generation newer).

What the Xeon workstation/server architecture provides is a fat I/O pipe that's still extremely important to music and video professionals who rely on near zero latency. Even photoshop cringes when working in print media scale designs that take up a couple gigabytes in a document.

Besides, Apple tends to move slowly when a product isn't front and center, usually milking a design for all it's worth. Tim Cook said they're working on a pro device that will be out this year and I believe him.

No bullocks. The Mac Pro introduced at WWDC last year was the most pathetic upgrade since the Mac Pro's debut. It's not just the CPU.

The "Pro" not only lacks Thunderbolt, it doesn't even have USB 3!

In fact the only port that my brand new retina Mac laptop and my Mac Pro have in common is the headphone jack.

I even need a new one -- but I just can't. I would feel like such a jerk if I bought the lame-ass Mac Pro that Apple's offering. (However, IIRC from his now-defunt podcast, after bitching just like I am doing, Arment eventually did just grit his teeth and buy one.)

The iMac isn't a developer machine. It looks on paper like a developer machine, but when put to the test of compiling large C/C++ projects day in day out it fails. We see more failures in iMacs than any other desktop we have at work. Way more hdd failure and gpu failure than our dells or other mac machines.

Why are you still running spinning metal drives on your developer machines?

The one reason people (like me) still want to keep the Pro alive, the one thing that no other Mac can do, is support for more than two screens.

Now, Apple already made it abundantly clear what they think of multi-screen desktops when they designed, and then failed to fix, the native full screen feature of OS X.

I love my Mac Pro mainly for this one reason, but honestly, like many other Pro customers at this point I'm just waiting for the other shoe to drop. I'm not going to have different platforms as my laptop and desktop, so if the Pro goes that's it for me and Apple - if by that time there is no real alternative to achieve the same thing. This is a hugely depressing thought. I hopped on board when OS X came out and never looked back to the Linux desktop since. For me, it has never been about the hardware, but OS X which always felt like home.

What platform would you move to then?

I'm not sure. Windows is actually looking pretty good these days, but I never really used it so I don't know what lurks beneath. I'd also give the Linux desktop another chance, but I'll probably run into the same old huge issues within the first few minutes (it's still my favorite server OS, though). There's not much else out there, that's why I said it's a depressing train of thought.

You can connect 3 external screens even to the cheapest 11" MacBook Air. You can do that with Retina MacBook Pros as well. I am not sure if you can with iMacs though.

The newest MBAs can power 2 external screens, as can the higher-end iMac (which isn't so bad since you can use the iMac itself as a 3rd). My "normal" 15 MBP can power one external (non-TB) screen, and I think the Retinas can power two TB + one HDMI. A lot of this assumes I'm buying Apple's expensive Thunderbolt displays, however.

http://arstechnica.com/apple/2012/06/new-macbooks-can-manage...

The Retina MBP 15" (not sure about the 13") can do two Thunderbolt displays or any two Displayport screens, plus HDMI.

Ah, OK, thanks.

Since this thread has re-opened a wound, I've been browsing external Thunderbolt PCIe cases. Putting graphics cards in these is probably also on option, as long as I'm not doing anything seriously taxing with them. But holy fuck, they're expensive as hell, too.

The Mac Pro lets you have up to six. That's more than I personally need, but those who need that many need them.

Would a KVM solution like Synergy address your needs? eg an iMac and MBP side by side sharing a keyboard and mouse?

I tried that once out of curiosity but it's a clunky solution. You can't really drag and drop properly and it's not possible to move apps/windows across. One solution I'm about to try out is using the large new iMac and two Thunderbolt displays for a total of three which would be optimal.

What's not optimal is the price of these things, however. I liked the fact that I could hook standard displays up to my Mac Pro, but the multi-Thunderbolt solution looks workable as long as one is willing to shell out a thousand USD/EUR per display.

Have you tried Teleport[1]? It's not perfect, but works very well.

[1] http://www.macupdate.com/app/mac/14042/teleport

It's the same as Synergy I think, with a nicer GUI though. I haven't tried this one specifically but it's not what I'm looking for. These Synergy-like apps are great, but it's something I'd use for attaching a random Windows or Linux box into my setup. It works well as a virtual KVM switch.

Thank you! Teleport just works. Great tip.

I had tried synergy a few times and never succeeded.

You can use non-Thunderbolt displays with your iMac. The DisplayPort adapters still work with the thunderbolt ports.

It's kind of ironic that Apple seems to be leaving the graphics professionals, which are about the only group to never abandon them during the tough times.

I couldn't agree less. A retina iMac wouldn't do it.

We'll probably see a new mid-size form factor aimed at professionals, slightly larger than the Mac Mini but with more horsepower and expandability. It'll be called "Mac" straight up. It'll be a throwback to the old Mac G4 Cube that designers and graphic artists loved, with marketing to match.

No reason to have a retina display integrated iMac—it's just not that exciting. What you'll see is the Mac Pro re-imagined for the modern world; one in which the existing Mac Pro makes no sense.

At least, we can hope.

The form factor will have to be significantly larger than any Mac Mini in order to accommodate a real workstation-class GPU. However, I wouldn't be surprised to see the PCIe x16 slot as the only expansion slot on the motherboard, with all the PCIe 2.0 lanes routed to Thunderbolt. That would enable the motherboard to be only slightly larger than mini-ITX for single-socket single-GPU systems.

Retina iMac is inevitable, do you really think we'll be staring at pixelated fonts on any product in the Apple Store in 3 or 4 years?

"A few power users will complain, but most won’t care: by that time, most former Mac Pro customers will have already switched away."

This is a statement that is general and without any basis at all. Where does the OP come off writing "most former Mac Pro customers" I can almost guarantee that this is simply pure conjecture with nothing to back it up.

"There’s even less of a reason to buy the Mac Pro today than I expected."

I run 3 monitors off a Mac Pro (24/30/24). I don't believe there is any way to do that on anything but a Mac Pro. I could run six with the two video cards I have now but can't fit them on my desktop. And I'm not using it for video editing or anything like that. I just like to work with as much screen space as possible.

The retina MBP can easily run 3 monitors, and more with a USB graphics card. I have 4 (24/27/24/23) running off mine right now.

Thanks. Which usb graphic card do you use and are you happy with it?

Maybe the answer is just to sell a version of OS X that is licensed to run on commodity X86 hardware, then just charge $1000 for it.

For 3D/VFX work you get so much more bang for the buck with a PC and Win/Linux these days. We're a mixed environment shop, but the Mac Pros lag so far behind the new PCs bought recently, for a fracture of the price. It's not even funny any more. Then there's the network - I don't know what it is, but all PCs in our network are about 50% faster than all the Macs. Personally I just use my MBP these days for some developing and personal stuff, that's it. Everything else has moved to PC entirely, and I can't complain. Apple would have to come out with an insanely compelling product to sway me back.

And why not just by a Mac mini and thunderbolt expansion chasis? achieves the same goal doesn't it?

e.g. http://eshop.macsales.com/shop/Thunderbolt/PCIe_Chassis/Merc... , http://www.magma.com/expressbox-3t , http://www.sonnettech.com/product//thunderbolt/index.html

It does not, because you are still stuck with i5/i7 processors, whereas workstations, especially those in high performance applications, would be much more suited to a Xeon.

I run a linux workstation with dual 6 core Xeon Sandy Bridge-EP's, with 64 GB of RAM. Being on an i7 would be limiting.

To put a needle fine point on what market is not being served here: content producers and editors who need all of the CPU, GPU, disk space, pixels, and I/O they can possibly get in a relatively price-insensitive business environment.

Homebrews, file serving, enterprise, SOHO, and those with nostalgia for DIY home desktop are non-factors now. The creators need a no-compromise beast with a flawless Adobe and Avid story.

If Apple won't provide it, they're throwing their crown jewel customers in the trash.

I see there being basically three types of person in this market niche:

- Those who want more storage without external latency

- Those who want more than one or two monitors

- Those who want addon cards.

Thunderbolt is capable of solving all of these problems, and in fact there's already market solutions for them. The problem is that the enclosures cost more than the Mac itself, and that's not including the hardware you want to put into them.

Thunderbolt drive enclosures and RAID systems are already starting to fall in price and will continue to fall as TB gains marketshare, but the PCIe cages are still STUPID expensive. If Apple were to release their own bus cage for people to add PCIe cards to their macs at a sane price point, suddenly the MacPro stops being the only Mac with that capability.

A chain-able TB display adapter would also go a LONG way to expanding the Mac's capabilities. One DisplayPort output per bus just doesn't cut it for some power users, and the only alternative is significantly slower USB adapters.

> A chain-able TB display adapter

You do know that you can already hook up more than one thunderbolt display to a single thunderbolt port using daisy-chaining, right?

And if I felt like spending $3000 to replace all my existing monitors, that might be a viable option.

Unfortunately a lot of thunderbolt hardware saves the extra cost of a tb-through port. So you're stuck with one or two devices.

I don't know if other thunderbolt displays (are there any yet?) support it, but Apple's do not support daisychaining a display through another display.

I have the perfect name for this product that would sit between the Mac Mini and the Mac Pro: The Mac Plus.

rimshot.

Not Mac MiniS ?

Don't you mean the Mac SemiPro?

I just want a dual or quad socket mac pro with haswell processors. I'm forced to use a Mac for iOS development, and large compile times just get annoying. Hackintoshes are not quite right at times and can screw up builds.

Are you saying Hackintoshes are known to sometimes generate executables that differ from those made by real Macs? That seems really unlikely to me.

As a person who used a hackintosh as a primary development machine, the problem is more that it is difficult to keep a hackintosh up to date, and due to Apple's aversion to backwards compatibility, this poses a significant problem for generating executables that will be accepted by a toolchain released 30 days ago.

If someone were to design a Mac Pro Mini, then it really needs both the memory and hard drives easily replaceable. I can even handle all flash if it mSATA and not some funky connector. It needs two (4 would be awesome) thunderbolts, usb3, and two (working, solid) ethernet connectors (one for network, one for storage). HDMI is nice (I would like an added input). I need 32GB of RAM on the minimum side. Core i7 and NVIDIA GPU would be fine. Include an iSCSI initiator.

If the next Mac Pro is any good, I predict a huge buy on its first day due to two years of pent up demand.

I wonder, since Apple is obviously abandoning this market, why is Microsoft trying so hard to alienate the power desktop user market with Windows 8 instead of aggressively trying to bring this group of Mac Pro users over to them. Of course they would have to improve Windows into a high-quality OS that works for them and find OEM's to partner with who are willing to find out what hardware these users want and provide it. Maybe it's a forlorn hope, but it's better than flailing after the tablet ship that sailed past them two years ago.

Because this strategy is making a shit-pile of money for Apple and Microsoft would rather chase this shit-pile of money rather than take Apple's table scraps?

That would make sense if they had any chance of getting that money with Windows 8.

I agree completely.

Lately everyone seems to be convinced that we're witnessing the death of the desktop computer, but I wonder if this could be partly due to the lack of options in this market. Apple and Microsoft have both dropped the ball here. Apple has eliminated hardware options that power users are looking for, while Microsoft still doesn't provide a decent POSIX environment or even a passable terminal.

I honestly wonder what the world would look like if Microsoft were to suddenly switch focus and deploy a version of windows with full POSIX support built in, a decent terminal, a first class package manager, and a decent app store. Instead they are removing features like the start button that confuse existing users, and adding a Metro UI that power users don't want.

Maybe I'm just getting old and cynical, but I think we would be witnessing stronger desktop sales if Apple and Microsoft were providing options that users actually wanted.

Powershell isn't a passable terminal?

No. It's a decent shell, but the terminal is still that horrifically slow 80x30 PoS.

Ah ok good point.

So, basically you want MS to make a Linux distribution?

I'm not sure if you're being serious or just sarcastic, but no - that's not at all what I'm suggesting.

OSX provides a POSIX environment, and OSX is obviously not a Linux distribution. In fact Microsoft already provides SUA (Subsystem for Unix Applications) which is basically a POSIX environment, but iirc SUA is only available in certain versions of Windows, and only as a download from Microsoft, post installation.

So no I'm not in anyway suggesting Microsoft re-architect Windows as a Linux distribution. I'm suggesting that Microsoft make POSIX tools an integral part of the OS, like they are in OSX. For example, making bash the default shell, providing cmdline tools like powershell cmdlets to manage the system from the command line, providing a useable terminal etc. Basically similar to what cygwin and msys provide today, but better and with deeper integration in the OS.

A Unix system intended primarily to run on X86? May as well be a Linux distribution. At that point it probably doesn't make economic sense to maintain your own kernel.

Unless, of course, you have these millions and millions of apps that need your on kernel to run.

Because innovation in this space is driven by the low end.

You can do about 75% of common computing tasks on a $30 ARM computer on a stick. The notion of a $3,000 PC is as relevant today as a $10,000 Sun workstation was in 2003.

Thing is, quite a few of those 75% things are created by the 25%.

Absolutely. 75% of the workstation market had shifted to Windows by 2003, and the remaining 25% were Sun/HP/SGI/etc.

But guess what? Cheap wins. The last vestige of the old Unix workstation market is the Mac Pro.

I have the rMBP with 16GB RAM and a giant SSD. Add the 27" Thunderbolt monitor and I think I have the new Mac Pro.

The power consumption is great and the performance is top-notch too.

But try rendering a large 3D or video sequence over night, and watch it melt through your desk.

Oh, and 16GB is nothing when working with e.g. After Effects.

It doesn't get that hot, even while running full-bore. I don't render video or graphics, but I do simulations and run a lot of VMs for various test modes. I suspect that fully loaded CPUs are going to hit a maximum temperature regardless of the application that is running. And yes ... I can't expand the RAM beyond that 16GB (it's actually the limiting factor for my VM count).

16GB of RAM is not a lot in workstation applications. I have simulations that use matrices which take 30GB of RAM. This is for one large matrix, where 24 threads are accessing different parts of it simultaneously. Your MBP is physically not capable of doing this.

I see it going the other way. Apple needs a bridge product between the iPad and the iMac. They are also getting clobbered by PCs in the business market. I suspect the big announcement will be a mid-range line of macs aimed for large scale roll-outs to corporations and governments. I read an article backing up this viewpoint in Business Week a while back, but alas it was only in the print edition.

I don't see that happening.

Apple products are luxury items with high margins. They don't need to or want to chase that mass PC money.

> Apple products are luxury items with high margins. They don't need to or want to chase that mass PC money.

This reality gets less true every day. It's the same argument people repeated to insist that the iPad Mini would never happen. Or the Mac Mini. Or $0-after-subsidy 8GB iPhones.

I don't think it would be surprising at all to see Apple look at business computing as money they no longer wish to leave on the table.

Oops, but why I always feel that even Mac mini has a substantial margin compared with its Windows counterparts?

Apple already sell a relatively cheap desktop computer in the mac mini. I don't think they could make anything much cheaper without digging into their margins.

Microsoft's value in the business market is their suite of server desktop & cloud products and the integration between those. Adding all of that stuff to OS X would be a big undertaking and would probably take away a lot of the simplicity of the Mac.

I doubt that will happen. The kinds of companies that are willing to give Macs to employees are issuing laptops. The big companies and governments that need desktops (think call centers, account, etc) are just not the market for Macs. They generally have a bunch of applications that only run on Windows, and the cost to switch is just too high.

Business has been switching away from desktop applications for years, outside of Office I don't see this as an issue.

Well, they just announced an i3 version iMac for education - they might extend that to large enterprise orders.

I wanted, MacPro in Cube Shape. Dual CPU ( Not Xeon Only, ) and up to 64GB of Memory, Dual Geforce GTX Titan, Dual Thunderbolt, SSD in Raid.

Because the Margin of SSD and Geforce GTX is much larger, Apple could make their own SSD and Buying Geforce GTX Titan at discount price. Making the New MacPro as similar pricing if you have built it up yourself.

I thought I had seen it speculated that Apple's "Made in the USA" efforts would be for the next Mac Pro.

Sounds like what is needed is a modern version of the SE/30.

Let's call it the iSE/30!

Reply on news.ycombinator.com