Apple Towers: Heading for the Long Goodbye
If you divide your time up betwixt writing revolutionary documents, serving as the third President of the United States and looking for Mammoths there is every chance you are Thomas Jefferson. Unfortunately, for President Jefferson, he never got to see one of the Mammoths he was so certain laid over the next rise because, by the time he got around to looking, all the beasts were long since extinct. No one is quite sure what happened to the Mammoths, some blame hunting pressure others cite climatic change as the likely reason. The question, if you’re a Mammuthus jeffersonii anyway, is moot. All the Mammoths are history. Soon to join the club of once semi common and suddenly gone is the PowerMac tower form factor. Unlike the mammoths in this case we won’t have to puzzle over the causes for the demise of the hulking aluminum enclosure; we are seeing the causes for the tower’s inevitable demise before our very eyes.
Why kill the tower? Well the reasons are bountiful. Let us begin by examining what the switch to Intel means for the Tower line. When Steve Jobs first confirmed the move (and it must be noted that it was the worst kept and least believed secret of the computing age) he went to great lengths to ensure the Mac community that the move was based on unassailable logic. The Intel chips, Mr. Jobs deftly noted, were so powerful that a new metric was going to be needed to measure their performance and that measurement would be performance per watt. Performance per watt is an excellent way to gage a notebooks computing acumen but tower users are a little less concerned about wattage and a little more concerned about raw performance. In fact many members of the tower clan would gladly run a small self-contained nuclear reactor solely to power their computer. Objections of the tower lovers aside we note that Steve Job’s (and later Intel’s) statements give us new insight into the thought processes of the industry giants: computers are fast enough so it is time to focus on making them run with a lot less juice. Without hot chips heating up the interior you don’t need room for multiple fans or cumbersome water-cooling hence the heft of the tower will no longer be needed for cooling consideration. The size of a tower can be significantly reduced and that is the first nail in the coffin for the PowerMac tower.
Of course if the tower is no longer produced a lot of expansion options go away as well. This will no doubt rankle many long time fans of self-upgrades. You know, the folks (like me) who like to slap an extra hard drive in or enjoy swapping out the DVD burner. Truthfully the necessity of these ministrations is becoming questionable. For all but the most demanding applications a FireWire 800 enabled external hard drive is more than adequate. The same holds true with DVD drives, with the transfer speed of FireWire an internal connection is no longer a must have feature. External drives also feature one bonus that internal drives lack: When it comes time to get rid of your computer the drives are trivially easy to keep. With the question of expansion addressed the coffin lid is getting pretty close to horizontal.
Another consideration in the whole “How long will tower last?” discussion is the big man himself: Steve Jobs. Steve has shown time and time again that he despises fans, large from factors and expandability. He, for example, insisted the Apple /// have no fan which led to the unfortunate experience of actually lifting the computer off the desk and dropping the machine in the hopes that the chips that were unseated by heat would be reseated by the sudden stop. For Steve’s distaste of expandability one need only look at the first Mac. Opening up the original 9” screened box of computing joy took a special wrench and a certain willingness to expose your self to a very unpleasant electric jolt. Mr. Job’s fixation with small form factors hardly needs explanation so it will suffice to note the ill-fated Mac Cube and surprisingly popular Mac Mini. You can be assured that if the opportunity presents itself the designers at Apple will be only too happy to drop the next generation of tower plans into the dumpster.
While all the above reasons are significant none of them, taken individually or as a whole, are compelling enough to warrant killing the tower Macs. As much as Steve Jobs may despise large computers if the machines are bringing in the green you would expect Apple to keep making the towers. Unfortunately the G5 towers have never sold well. When they were first released supply was constrained so the initial lackluster sales were understandable but when supply finally caught up with demand it became apparent that there just wasn’t that much demand. It may have been that people were loath to spend such a large amount of money on what is a better computing experience but not a revolutionary computing experience. More likely there was just no compelling reason to buy the PowerMac G5. Professional applications like Photoshop did not require a G5 (though there are certain high end movie production apps that demand a G5) and Mac OS seemingly gets faster with every iteration. Why buy a three thousand dollar tower for a performance boost when you can just wait for the next edition of Mac OS X?
So it seems unlikely that the tower form factor will stick around when the transition to Intel is complete. Sure the people who buy more computer than they need to make up for perceived inadequacies will be disappointed (note: you can’t buy a golf swing or computer skills) but the majority of folks will embrace the tower free future.
del.icio.us
Digg
Facebook
Slashdot
StumbleUpon
Comments
Chris this article is somewhat incomplete. If the tower goes the way of the dodo then what replaces it?
My opinion on the subject is that towers are indeed going to go away and the network will then reign supreme. In 10 yrs the desktop will simply contain a monitor and speakers and network connections.
The “guts” of the computer will reside in a locked box in another room or the basement. Here is where the CPUs, Video cards and storage devices will remain. You won’t have a dedicated drive to your computer you will uses NAS/SAN type storage systems. The multicore processors and hardware will support virtualization down to the I/O. Power will be dynamic…when everyone is utilizing the hardware performance will be distributed as fit. Meaning the parents may get priority scheduling over children. This is all adjustable.
You will be able to boot (Net Boot) from different LUN(Logical Unit Number) on the storage pool which will be sliced into differing RAID levels. Your GPU will be multicore and in SLI configurations with enough oomph to run gaming for the entire family or superior pixel processing for one perso.
Because the noise and heat making devices are in a basement or noise sheltered closet they can be designed for robustness. Fans, big black chassis and other unsightly things won’t matter when they are out of sight out of mind. Want to add more hd…simply plug another drive into the chassis. Want faster video? Plug in another or faster GPU.
The scalability and manageability of this scenario is tangible. Homeplug networking will be mature as well as faster wireless. The Network is King.
I think it’s probably safe to say that in the Windows world, users are more apt to open their box and upgrade it somehow than Mac users. Recently I suggested a friend get a second hard drive for his his PC. His first thought was to an internal one; mine was a USB 2.0 that he could plug into it…
That being said, I know a lot of Windows people who absolutely drool over the internal design of Apple’s towers.
Now that Apple will be making systems that users can buy and also install Windows on, won’t they want to go after that market too? In a way that might be an argument that Mac towers aren’t going anywhere.
The tower is not dead yet. Recall that one of the largest segments for Mac users is still audio/video multimedia production. On the high end, these machines still require PCI (or PCI-X) cards. Some of these solutions are moving toward Firewire, but for the time being there still exists the need for the open PCI slot or two for this market segment.
Also, there are still applications for which dual/wide ultra SCSI drives are better than IDE (keep in mind that most Firewire drives are just IDE drives with a Firewire controller) as well as people who need a hardware RAID controller who aren’t looking to buy an Xserve.
While the tower isn’t the largest seller in the Macintosh line, it is still an important form factor for a significant portion of the Macintosh market.
You didn’t really address the upgradeability factor at all. All of your arguments apply at any time in the last 15-20 years of computer history. It has always been possibly to get equivalent speed from external devices as from internal ... high-speed SCSI card, anyone? Etc., etc… The reason people like an easily upgradeable box has nothing to do with that, and never did, so your narrow counterargument is completely irrelevant.
Simply put, people who like to expand internally like to do so because (a) for obvious reasons they want to keep the physical collection of satellite devices to a minimum, (b) it’s simply useless and wasteful to keep powering internal devices which have been superseded and will no longer be used, and (c) it’s cheaper, usually by more than half (duh!). Yes, there are some advantages to getting an external device but absolutely nothing has changed about this trade-off today vs. ten years ago. On any computer equipped with the high-speed interfaces of the day (i.e. FireWire or SCSI), speed of interface has *never* a particularly significant reason for going with an internal device. (In fact, with some flavours of SCSI that were popular in towers, the truth has been quite the opposite.) Therefore, there will always be a big box (big enough for additionnal bays) and your theory is plainly bunk.
Ten years ago, you needed a top of the line computer to get any kind of decent performance out of Windows or Mac OS. So whether you were doing intense database/spreadsheet stuff or just a word processor document, you needed an expandable tower/box.
Nowadays, though, a lot of home and small business users do mostly word processing, e-mail, and web surfing. Maybe a spreadsheet or two here and there.
For that kind of use, a G5 iMac is plenty of horsepower. For what home office & small business people do nowadays, you don’t need a tower. And so you make a very good point when you address these users, Chris.
However, you don’t address the Pro market at all, as some others here have alluded. If you’re running DVD Studio Pro or Final Cut Pro, a G5 iMac isn’t gonna cut it.
So, while the tower market is significantly smaller than it ever was, it still must be very profitable—but even higher profits on a very small market won’t encourage a lot of investment on Apple’s part. So Apple might spend less time & effort designing the ultimate tower and have something more generic looking. Either that, or Apple will migrate power users to dual-cpu XServes or something.
But there’s definitely a section of Apple users that need RAID and expandability beyond a Firewire 400 drive or two that you get with iMacs or Mac minis.
Computers are not fast enough today for all applications, not even towers. It’s still slow to work with video or Hi-def images no matter how advanced the system is. So IMO towers will still be around for a while. What would change this would probably be a relatively revolutionary new technology that makes ALL computers smaller so that, say, the most powerful computers would be the size of an iMac (the New Technology would make it upgradeable enough) and the smaller ones maybe somewhere between todays Minis and iPods.
My guess is HD and Ram would be somehow combined (can solid state achieve this?), chips would become smaller and capable of handling video and audio, and portable storage would go the way of mini-DVDs or memory cards/flash drives. Maybe the evoulution of the Net could in the future eliminate the need for portable storage altogehter.
Any thoughts?
You leap from conclusion to conclusion with such ease and grace I’ll lay $100 right now that you could jump the Grand Canyon in your bare feet.
By your logic, and I chuckle meanly when I use the word logic, Apple would have killed off the tower already. That the tower doesn’t sell well is totally irrelevant - it is selling to a market segment that needs it and the iMac is selling to a market that doesn’t need it. Those of us who own towers NEED a tower. Chaining 4 drives with firewire isn’t a solution. End of story. PCI expansion is mandatory. End of story. Tower. No tower I’m out of business. End of story.
In fact, I’ll lay another $100 down that once the professional desktop line has been moved to Intel processors the new Intower will have MORE expandability than the current G5 tower. Why? Because its thermal signature and cooling needs will allow it. My G5 tower isn’t just bigger than my G4 tower, it last LESS expandability due to the fans and carefully crafted airflow.
Interesting comments all around. For the record I suspect the future will see the iBook/PowerBook, iMac, mini and some slightly larger cubes. For those of you can’t imagine life without a tower, and it seems a strange thought indeed, I encourage you to investigate what kind of computers Steve when unfettered. You’ll find no towers but you will find cubes and small stations (NeXT cube and NeXT station)
My hope is that Mac x86 towers will be far more expandable than current ones. There are some “killer workstation” trends right now, such as GPU cards working in parallel, internal hardware RAID, quad CPU motherboards and so, that PC boxes seem to be able to hold without that much effort comparing to G5 boxes.
I wonder if Apple would make a stab at doing an SFF-sized box, sort of a MacMini Pro: an as fast as possible box with expansion capabilities similar to Shuttle’s ones. Most pro users would do great with such a thing.
You’ll find no towers but you will find cubes and small stations (NeXT cube and NeXT station)
Chris,
I think comparing the state of the industry now to then is a little difficult. Remember that NeXT as a hardware company was a failure, which is why they switched to software and services.
You also fail to remember those periods of Mac development when multi-slot machines were discontinued (IIfx, Quadra 950 / 9500) and Apple was slow to replace them (sometimes it was a year before they made a new tower). Most Mac CPUs had 3 slot set up (2 of which were often occupied by network and video cards) and I remember reading many “We need more expansion!” articles and op/ed pieces in the trade magazines during that time.
Eventually, video and ethernet was moved off NuBus/PCI and onto built-in ports so that there were more free slots, but the accepted standard became a minimum of 3 open slots for pro-level users
Those pro-level users are still around, and even though Steve Jobs likes design simplicity in smaller form factors, he is still answerable to the board of directors and the shareholders. Even he isn’t ignorant enough to cut off a significant portion of the Mac market. Since his recent mantra is “Grow Mac Marketshare”, making major business decisions counter to that philosophy would be unwise and I believe unlikely.
Additionally, even though the tower lines doesn’t sell as many units as the others, the profit margins on the pro machines are much better than the consumer margins.
Multimedia users kept Apple viable even during the deep, dark years. Abandoning those users now would be a terrible decision from both a PR and profit margin standpoint.
You can’t abandon a tower, not on today’s technology. When you can provide dual-graphics card level GPU performance in a Mac Mini sized enclosure and get 500 GB solod-state chip sets to replace hard drives, then maybe the tower can go.
As it stands, they just need a tower with more PCI-X in it. PCI-X 16 for the graphics card, and make that two slots for linking them. Give us a dual-proc, dual-core machine and gigabit ethernet, fast as heck RAM, and SATA2. Give as 2 FW800 AND 2 FW400 ports, and 4 or so USB 2.0 ports. Make all the wireless anything stuff built in for those who need it. This is the stuff that really high-end work needs. Serious, pro video editing. 3D work. Serious, pro audio work.
I’m tired.
While there may be external devices for pretty much anything an end-user consumer will ever need nowadays (this totally excludes the pro market), people such as myself will buy towers for one reason: self-containment. I like the fact that I can fit everything I need in *one box*. This makes moving the box, and positioning everything around it, much much easier. Could you imagine having 500 GB of external storage shared between two FW drives? That’s at least two other satellite devices, excluding printer, mouse, monitor, speakers, cable modem, whatever.
With a tower, it’s all there in one package. I do wish Apple would bring back the inverted-plug on the back of towers (as my 8500 had) so I could plug the monitor directly into the tower; this would allow me to get away with only two plugs on my desktop (tower/monitor and printer), instead of having to snake a power bar up there as it is for three plugs. (And that’s only the bare essentials; excluding add-on’s like a USB hub, ethernet hub, speakers, second printer, etc!). I don’t need all of that desktop clutter.
The tower is here to stay.
I haven’t even bothered yet to read the other comments. I imagine many of them, from sensible people, will already have covered my two cents. I registered out of shock over this article. Can the writer actually be serious? Why in the H would you want a small computer with a thousand things plugged into it? Nevermind that the tower is hardly Apple’s invention. Nevermind that every other computer hardware manufacturer makes towers. I could keep going with this “Nevermind” motif for hours. Good lord I hope the writer is just being flippant because otherwise he/she is writing for a computer column and seems to have no real grasp of the various roles a tower computer has played in the past and continues to play in the real world. I have a G3 B & W that has been upgraded to pretty respectable performance by today’s standards THANKS to the fact it’s a TOWER. I have a G3 iBook that is technically newer than the B & W, but is and forever will be stuck in 1999 performance, because it is only expandable with externals options.
Good lord dude, are you really that out of touch?
In the future we won’t have central heating or air since we could just put pint-size AC or heating units spaced out every five feet in our homes. THAT’S how stupid this article is.