It's either a lie, or an ex-Apple employee posting. Not an ex-Apple employee now, but very very soon
More likely a lie.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
2011 iPro = Mac Pro replacement
- Thread starter tpavur
- Start date
- Sort by reaction score
It's either a lie, or an ex-Apple employee posting. Not an ex-Apple employee now, but very very soon
More likely a lie.
Thats a type one class Macally external hard drive case made to look like a minature mac pro.
Same here, there's not much that I can think of adding that would really warrant a case redesign. Modifications . . . sure, redesign . . . No.
Since Apple is dramatically cutting back on their enterprise/high-end offerings to focus entirely on consumer/portable devices (understandable since that's where all the profit is at; something like 75% iPhone/iPad, from what's left 20% are portables and 5% or less are iMac/Mac Mini, with keyboards, random crap like the magical and revolutionary battery charger, etc, etc, etc, following that, with the Mac Pro limping along at the very tail-end of everything)... if Apple is not yet jettisoning the high-end/enterprise/pro users, altogether, then they really need something which is rackable and somehow at least fits into a 3U form-factor, instead of the enormous, giant, 60lb, bullet-proof slab of heavy metal the current Mac Pro is...
The current Mac Pro just takes up an enormous amount of space, the case was originally designed to house the PowerMac along with water-cooling, while it's beautiful industrial design, it's not too terribly effective at utilizing available space, most especially when it needs to fit into industry-standard racks (which are typically 1U, 2U, 3U, etc...) The current form-factor is a major headache if you want to deal with, for instance, 3-4+ Mac Pros running headless.
There is no more Xserve, and the current Mac Pro market is minuscule in comparison to Apple's other highly-profitable lines. Something reasonable which can absorb the Xserve & Mac Pro market, house a few SSDs, and dump the optical drives which are nearly useless (seriously, who cares? I long ago pulled the optical out of my personal Mac Pro to make room for more drives, if you need an optical, well ... buy one in an external enclosure, they cost $90 bucks, and are so slow that just plugging it into a FW800 port is more than acceptable speed. Out of all our work machines, I believe there is only one Mac Pro left that still hasn't had the optical drive tossed out to make room for something else).
Unfortunately, the 3U (5.25") format poses serious challenges for cooling when used as a desktop/pedestal, as there's no HVAC/additional forced air cooling system in addition to what's in the computer enclosure as is usually the case with racks (relevant to DP systems in particular).
5U (8.75") would be sufficient, but that's what it is now IIRC. Just add the ability to remove the handles and install rack mount brackets for those that need a rack installation and that's taken care of. As it stands however, the market seems thin, as Apple's never been big on enterprise support/solutions (hardware and software offerings, tech support, and warranty), and even less so in the server space.
It's not perfect, but the additional space is needed for cooling. Particularly as they've gotten rid of liquid cooling, and gone back to heat pipes and forced air solutions (cheaper, but it tends to require additional volume to do properly).
Just a simple mod mentioned above would do the trick.
But there's something else to consider, and that's a 3rd party solution + Mini's.
The XServer variant (designed for the TB port models), and a pair of Quad core Mini's, and you've got a lower cost/performance ratio than the MP or XServe could ever produce.
Just add fast networking via the TB interface (which would need to be done with the MP's as well, assuming network throughputs will exceed 1G Ethernet, which isn't hard to do on a server).
There's a lot of serious downsides to this. No redundant power supplies. Far less hard drive bays. No hot swappable hard drives. No fiber channel. No user service kits.
Yes, the price ratio is good, but for a lot of places, this is a total non starter. You've got to meet minimum requirements to compete. I've seen server admins grudgingly give up the hot swappable hard drive bays, and the redundant power supplies, but it's been hard. I can't see those same people moving to Minis.
I get your point, and in a proper enterprise environment, I agree without the slightest hesitation. The support issues and lack of integrated enterprise solutions are another major reason to skip Apple IMO.
But keep the following in mind with Apple's products:
- The MP doesn't offer redundant PSU's (or any other Apple computer currently on the market).
- Storage can be handled by an external solution, such as a DAS, NAS, or SAN via the PCIe slot that's in the XServer version for TB based Mini's (takes care of the HDD bay issue for the most part, though the argument is still relevant for internal OS disk/s - they can be moved externally as well, but it's clunky <USB/FW/2nd TB port solutions>, save on a RAID card).
- Support sucks for the enterprise market (they even treated the XServe as a consumer system; only able to upgrade to a 3 year warranty was a joke). Since they don't offer a full line of hardware and software solutions, they can't provide purchasers with a single point of contact in a number of cases either.
In the case of FC and the Mini solution:
Take a closer look at the RacMac Mini XServer. It connects one of the TB ports to a single PCIe slot located in the rear of the 1U enclosure, so you can install things like an FC card (or RAID, 10G Ethernet, Infiniband,... that supports OS X).
I'm not saying this is ideal, but it's not quite as bad as you make it out to be. And for low cost solutions (also as a means of further cost cutting), the market will find them acceptable. Particularly in the independent to SMB portion of the enterprise solutions market to begin with, where budgets tend to be tight.
It's also a major direction the market's headed, including running a cluster of ARM or Atom based systems with graphics cards used for GPGPU use as a means of increasing performance (seriously, this isn't a joke).
And I'd be amazed if these sorts of solutions will run redundant PSU's , or add much in the way of internal access, though I do expect more than the Mini from other vendors that have a significant presence in the enterprise market (at least be able to swap out the OS disk, and better support and integration solutions ).Right... but typically external storage in enterprise is accomplished on large RAIDs via fiber channel, which goes back to the Mini's lack of fiber channel.
If Thunderbolt gets faster, one could probably do fiber channel via Thunderbolt, but that's hardly a sure thing.
The inability to repair Minis on site is also a huge pain. How exactly do you swap a dead drive in a Mini easily? With an XServe, you just pull the drive from bay 1, put a new one in, and you're done (with the Pro requiring you to crack the case.)
Apple didn't sell many XServes, and it was buggy, so I understand why they cut it. But the reason the XServe was released in the first place was to scratch an itch that other Apple products didn't.
Again, take a closer look... it can be use with a Fibre Channel card due to the single PCIe slot located in the enclosure itself (which is connected to the TB port).
Granted, it's only a 4x lane slot, but TB is still good for 800MB/s, and some real world testing has shown it a bit faster (~850MB/s IIRC on a Promise Pegasus R6 in a stripe set configuration). Depending on the FC card and implementation, it may not reach full speed, but it's close (i.e. 2 port 4Gb/s card would throttle a bit). But it's ~8x faster than the 1G Ethernet port at least...
Like I said, it's not ideal (may be able to crack the Mini case as well for a drive swap, though it won't be as quick and easy).
But most would be willing to deal with it it means getting the performance they need within their budgets.
And in some cases, even the internal disk issue can be gotten around, such as using a bootable RAID card (either a Pass Through or RAID 1 on the card for the OS/server applications).
But I suspect there's one thing you haven't considered; You can swap out the entire Mini as easily as a disk is swapped in an HDD backplane (keep a spare Mini on hand for just such occasions, just as you do with spare drives when thinking in large enterprise terms). More time would be spent updating the OS/server disk than swapping out the Mini.
There are other reasons as well, such as the poor support and warranty,and lack of integrated solutions that further reduced it's sales volume. Add in the fact users are no longer stuck exclusively to OS X for server needs (PPC gone, significant improvements in Win and Linux server applications offered, and it's easy to network to such a server/cluster/cloud), there wasn't really anything offered by the XServe that made it a *must have*. I suspect existing infrastructure was the biggest selling point, which doesn't generate enough sales to keep a product line going.
Ahhhh. Yeah, you're right, this makes the Mini a little more acceptable, although it still has issues.
Which is true. As long as the Mini doesn't change form factors or port locations.
The lack of good phone support for the XServe, or a good on site repair solution certainly doomed the XServe. I'm sure Apple knew that, and decided that the cost of spinning up the sort of support and AppleCare needed for the XServe wasn't worth it.
Definitely has merit for some applications.Ahhhh. Yeah, you're right, this makes the Mini a little more acceptable, although it still has issues.
I don't expect it will, as it would be based on a standard board format (suspect mini-ITX format as it's 170 mm x 170 mm, given the current Mini is 197 mm x 197mm).Which is true. As long as the Mini doesn't change form factors or port locations.
I've no doubt that money was the ultimate deciding factor. The ROI wasn't good, and investing more money would take too long before they'd see any return (would have had a battle to gain marketshare in the enterprise segment). Particularly compared to what they're getting out of the consumer segments they participate in.
Of course looking at how the iTunes Match and iCloud launch/performance has been, it might have been worth it to invest in in-house expertise. Outsourcing to Microsoft has not gone well so far. It almost would have been better for Apple to get OS X Server more stable, and throw a bunch of XServes into the NC facility.
I agree.
But the blame can't be laid on Microsoft in this case, as the development strategy was Apple's decision.
As per OS X Server running on XServes, add the development time to fix OS X Server to that of the iTunes Match and iCloud servers, and I suspect they decided to pass on it due to the total development costs.
I don't mind external optical drives, room for ssd and standard hard drives would be nice. At this point, i just want something with the Sandy Bridge Xeon
Well, the better question is if fixing OS X Server would be equal to the time of fixing Windows Azure running iCloud.
Given the number of developers that work for MS, Azure has a chance of beating OS X Server if it were a contest. Another fact to consider, is their entire market strategy for Cloud services is riding on it, so they're highly motivated as well.Well, the better question is if fixing OS X Server would be equal to the time of fixing Windows Azure running iCloud.
Nanofrog, I have a tremendous amount of respect for the sheer breadth of knowledge you drop here on a regular basis, and I can't find fault with anything you're saying.
But...
I'll bet you $100 that 2010 = The End of Heavy Metal in its' current form. What gets released as the next "Mac Pro" will absolutely not be the current form factor. It has had one of the longest runs in design history, but looking into my crystal ball, it's over.
It would come down to economy of scale (sales volume) of the MP, and if they're going to stick with it as it currently exists (workstation market is in a transitional state due to the increasing core count per CPU).I'll bet you $100 that 2010 = The End of Heavy Metal in its' current form. What gets released as the next "Mac Pro" will absolutely not be the current form factor. It has had one of the longest runs in design history, but looking into my crystal ball, it's over.
Which means 8 cores on a single CPU isn't too far away. Once this happens, a lot of users currently requiring DP systems will be able to shift over to SP systems running on less expensive CPU's (no longer require a socket capable of DP configurations). This is when I would expect a significant case redesign, as it can also become the "headless X Mac with slots" that people have been begging over for years.
Given this fact and Intel's roadmap, that puts it at Haswell. Which is why I expect the MP in it's current form will make it through Sandy Bridge and Ivy Bridge (cheaper to stick with the existing case with a mod or two to accommodate the new internals).
Agree. It most likely will stick around at least until Apple kills it's Server OS. MacMini's just don't have the storage options and you have to serve on something. XSAN, FC Server won't be phased out for a few years more. All signs that Apple is pulling up the enterprise HW stakes. They'll just work on SW integration.
Workstation market depends on other factors and I think Apple still likes having a certain presence in major content creation. They should give us a Black Anodized Case option though to keep up with the product look these days.