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.
Let's predict how the new Mac Pro will do
- Thread starter cnev3
- Start date
- Sort by reaction score
IMO, less significant than you might think.
First, the basic Mac Pro case has been around essentially since the G5 (2003) and has had its tooling costs/etc long since amortized. If we use retail prices, since one can buy a full ATX case for $100, this is a halfway decent guess for how much cost savings we might see from Apple.
Next, I'm not going to go track down how much a SATA controller chip costs, but a quick retail search on PCI cards for SATA finds 2 ports for $20, so call it $50 for the five that would be eliminated.
Next, call it $50 for the deleted Superdrive.
Finally, lost PCIe slots versus more TB chips, etc .. call it a push.
Added up ... you've saved a whopping $200 (retail).
IMO, I don't really think so. First off, a family friend who works at Dell told me (while we were both at a funeral last year) that for Dell, they were only saving $25/unit on paper. Even if we double this for Apple, that's still not really that much money, particularly on a low volume product. As such, the "Made In USA" is more for Product Placement & Marketing.
This looks pretty reasonable to me, although I'd bet that the likely Apple price is going to be $1999 starting - - from a very cynical perspective, it looks to me like the average "Tax" on a TB external peripheral (vs an internal HDD) is around $150, so just the four deleted 3.5" bays represent a $600 magnitude change - - simple math here could suggest a $1899 starting.
However, I'm also of the opinion that if Apple really wants to crack open the market and move units - particularly amongst the cynics who see precious internal expansion capability going away - Apple really needs to get it down to $1500.
If nothing else, this price point might help to finally shut up the whiners who still can't admit that the barnsale MSRP of the last-of-the-G4's from 2003 was just a one-time anomoly.
-hh
Just how much of a bath do you think Apple will be willing to take on these? $1999 wouldn't cover 1GPU, much less everything else in the can.
$1999 + replacing that internal storage. In my case, I'd need at least 2 external drive enclosures, just to hold the drives I already have ($1,000). Which will get me a bottom of the line Mac Pro.
TCO fail, any way you want to cut it.
Because they don't have to cost $1,999
A FirePro W7000 costs $699 retail.
A FirePro W8000 costs $1,499 retail.
If Apple wrangles 15% off of those prices then $595 and $1,275. So a pair being ~ $1,200 and ~ $2,550 (AMD's prices are greatly inflated to cover seperate speclized drivers. not needed on Mac OS X and product segmentation also not particularly needed on Mac OS X ).
1,200 (GPU) + $300 ( CPU ) + $300-400 (case + infrastructure ) + $100 (RAM ) + $250 ( SSD) = ~$2,199-2,299.
At similar kinds of bandwidth really just need one that holds all possibly 6 drives. What need though is a JBOD box; not a RAID box. The $1,000 is in part because adding a RAID controller cards to the mix... it is just embedded in the TB box. Don't really need one, let alone two.
The big price difference in the Mac Pro case versus generic case in the the materials. Miltiary grade aluminum versus generic cheap stamped alumnim and plastic. That actually doesn't amortize over time.
SATA controller effectively costs zero. It is in the chipset. It is even in the chipset of the new Mac Pro ... going basically entirely unused.
Throw in the HDD and Superdrive and probably not covering the SSD.
It is not likely to be inside or even at the edge of the Mac Pro range with two GPUs and about 8GB GDDR5 VRAM standard.
I don't think Apple is going to work backwards from Mac Pro + TB devices prices.
The bulk of folks buying this will be those whose storage is outside the box now anyway.
a. groups on SAN/NAS set ups.
b. groups that depend heavily on sneaker-net (swapping external drives ). [ Often "poor man's" version of group a. ]
c. users for which 256-500GB is enough for single machine storage. (e.g., developers, financial analytics , etc. ) Long term back-ups and archivial storage out of the box.
If reasonably priced JBOD boxes show up then the market gets bigger, but that isn't the core price driver.
On of the problems is that the 27" iMac screen isn't that huge of a cost factor. Push upper end BTO CPU and GPU into an iMac while ripping out performance of the Mac Pro to limbo into the iMac zone and going to run into a buzzsaw with folks who are not caught up in form over function ( "I hate integrated displays so iMac is not an option." ).
You just aren't getting it.
I am glad you know what my workflow needs better than I do. I need cores, followed by memory, followed by vast expanses of HD space in that order.
My boot disk is 2 256 SSDs in a Raid zero (and it is nearly full of programs). My data drive has 3.5Tb of Data, my iTunes library has 3.5Tb of data, JBOD my ass.
Read my sig - 2 SSDs (2.5" form factor), 4 SATAs (3.5" form factor) in 3 RAID 0s, with a separate 8Tb backup in a separate FW800 enclosure, - I'll need 2 boxes for HDs and a host of adapters. $1,000 for TB1 inclosures, not TB2, because it is still vaporware at this point.
At the end of the day, $3,000 for a 4/6 core configuration that maxes out at 64Gb. And speaking of RAM, What ever is shipped won't be enough, so those will be tossed to make room for at least 24Gb of ram.
TCO Fail.
I said absolutely nothing about your work flow. You made a comment about minimal GPU costs. Now you want to inject a misdirection into your workflow.
Two SDDs in RAID 0 is likely highly unnecessary if have a PCI-e based SSD.
This kind of thing along with short stroking HDDs is one of the things that drive up internal drive bay counts were are old work-arounds for problems that have better solutions now. The mindset of "this is the only solution that will work" over time tends leads to dead-ends.
That's nice. And exactly where did I say someone didn't have filled up internal drives. Oh that's right. No where!
Hah, you don't need 64GB of ram for an iTunes library. Also, your iTunes library is probably way too big.
Also, just buy 2 Mac Pro's. You can network them. Thunderbolt over IP.
Also, Thunderbolt isn't vaporware. It might be a little expensive, but if you're buying a Mac Pro because you require it, you should already realize this and factor it in.
You and deconstruct60 don't seem to understand the concept of TCO. Those of us who already own MacPros aren't just going to throw out everything and start over. It isn't cost effective. Of course, if you are coming from an iMac or Mac Mini, why not?
I don't have 2 SSDs in a Raid 0 because it saturates my bus (although that is nice), but because I filled up the first one, and it was cheaper to buy another 240 SSD and raid it with the one I already had than to purchase a 480Gb SSD.
It was cheaper to raid 2x2TB devices than to replace working devices with a 4TB drive. Faster thru-put was a bonus.
Form factors drive me to needing 2 external HD enclosures, unless of course, I want to add even more adapters - which will lead to yet another potential point of failure.
My workflow isn't iTunes, it is Z-brush/3DS Max (in a VM) - Poser (with a god-awful number of add-ons) - LuxRender and/or Vue.
They need all the cores and memory I can throw at it, because the software has finally become multi-processor aware on the OSX platform.
I am running a 5 year life cycle (I got my Mac Pro at the very end of 2007.) I won't be able to move anything from the 1st gen TMP to the 2nd gen down the road. Which is another strike against the TMP.
Thunderbolt 2 looks like vaporware to me. YMMV, but until I see products I can put money down on, it's vaporware.
I am factoring in the cost of TB enclosures with a TMP - which is why a Dell workstation is probably in my future. $4,400 gets me 16 cores, today, and I don't have to buy additional products to keep the same functionality.
This child-like belief that Apple is going to massively cut the price of their top of the line products is nothing more than wishful thinking, at best.
As far as my iTunes library, I am sure that you would be happy with the sound of 128k MP3s on your Bose stereo - I am not - when you have a good A/V system, you can hear the difference between them and loss-less compression.
3,719 albums, 48,078 songs isn't a large music library. I may not have top-billing in the movie of my life, but I'll bet my soundtrack is better than yours. Then there are the 1,000+ movies & TV shows.
You may be happy with what in economic terms are "substitute goods", but life is too damned short to live with substitute goods.
Oh jeez... another dude who thinks he can tell the difference between FLAC and V0, and thinks anyone who listens to a 128k rip is inferior. Well, I know what kind of person you are...
You're complaining about the new Mac Pro, when you're using one from 2007? Give me a break. If your 2007 machine can cut it, and you're not suffering enough to upgrade, I think you'll be fine.
You're talking about spending almost 5k on a PC, but you can't afford a $1-2K thunderbolt enclosure because it doesn't exist. Hmm... You don't add up.
I'll go slower for you.....
Who said I couldn't afford the TMP? I can. I don't give a damn about form - I am all about function. I am that "truck driver" Jobs was talking about.
The TMP doesn't appear at this point in time to be a good value in comparison between it and a Dell workstation.
Yes, I am using a Mac Pro from 2007. Why would I upgrade the hardware, if the software can't take advantage of it? Most of my software I use didn't go 64-bit & become multi-processor aware on the OSX platform until late 2011.
Now, the 1,1 doesn't cut it - my software is now completely 64-bit and will use all cores available. And then there is that life-cycle thing.....
I am at the end of life cycle of my current computer, so do I go with the TMP or do I jump platforms (again)? I will be spending about $5k on my next computer - the issue is that right now, I would get more computing power with a Dell for that price than a TMP.
The TMP will max out at 12 cores. I can get 16 cores with a Dell right now. 8 Ram slots vs 4 Ram slots means more expandability at a lower cost. A second GPU won't add any value to my workflow, although it may for others.
No additional desk space taken vs spaghetti junction with a couple of 3rd party boxes scattered across a shelf. I consider those 2nd & 3rd order effects, even if you don't.
As far as music.....
If the music you listen to is heavily compressed when it is recorded, then yeah, you won't tell much of a difference between an mp3 and a flac. If it is recorded properly, with plenty of dynamic range, you will.
I listen to a lot of classical and baroque music as well as rock. My A/V system has a pair of Maggies. Once you have had planar speakers, cones don't cut it anymore. That is just how it is.
What is TMP? Isn't it redundant to say the the Mac Pro?
What software are you using? Just curious. I'm guessing audio.
$5k is a lot of cash to throw down. I will say a couple things. You're almost always going to get more for your cash with some PC. The beauty is in OS X optimized for the hardware. So not sure exactly what your complaint is there.
If you don't need OS X, then I don't understand why you don't get a Dell in the first place. Mac hardware isn't really anything special. PC's use the same (similar) hardware.
It depends what kind of work you do, but you're always going to have cables sticking out.
I'm automatically skeptical of anyone claiming to be an audiophile. Just the way it is. I didn't say there is no difference between uncompressed or compressed, just that I am skeptical of those who claim to be able to tell the difference. All good since we don't know each other and it's a pointless discussion anyway.
TMP = Tubular Mac Pro....
Software - I do 3d rendering - so for modeling Hexagon, 3dsMax (in a VM), and Z-Brush - basic scenes are set up in Poser 2014, and then sent to Lux via the Reality 3 plug-in, or Vue for landscape scenes - then into Photoshop Elements, and finally into Comic Book Creator.
The issue was timing. Most of this software didn't go 64-bit and/or multiprocessor until late 2011. And then there was the lackluster 2012 "update". Since I wasn't looking to update until 2013, I could wait and see what Intel's road map looked like.
$5k is a lot to throw down in 1 shot, but on a 5 year life cycle, it is a better deal than going the iMac route.
If I had bought an iMac, I would have started with an iMac 7,1, then moved to an iMac 9,1 in 2010, then to a 12,2 in 2012 - the cost would have been higher, with less performance - assuming that I didn't bake one - (Rendering is hard on sealed systems like the iMac, because Sir Idiot Boy doesn't take heat dissipation seriously - I have multiple Ives designed products that cooked themselves.) Maxing the ram out, plus all of the extra storage costs would have made it painful.
And the whole time, the hardware would have been the limiting factor. With the MP, even if the software could only access 2Gb per process, I had enough ram that I could just switch to another program while another one was cranking on something, so I wasn't seeing a slowdown in whatever I was doing. I can't have multiple applications running in 8Gb of memory.
In the workstation segment, there wasn't much difference in the way of costs at the time. I have a Dell Precision 690 on the other side of my desk. It was priced out the same as a 1,1 MP on release - I paid $300 for it when it came off lease (2008).
I was already in the OSX ecosystem when I discovered Poser 5. I grew into it, so making a clean cut earlier didn't make any sense.
As far as the cables, every port I have is in use - the TMP adds to that.
And I see we both agree that those audiophiles are a bunch of pretentious sphincter muscles.
Someone's not happy about losing customers in their target market
I wouldn't be making an assumption on a mass migration to MP 6,1.....
There are still lots of 1,1s, 2,1s, 3,1s 4,1s and 5,1 that could use new cards.
Oh, I don't really believe that Apple will be willing to not leverage a fat profit margin out of this product line ... but that's not really what I was saying. What I was saying is what the price point *needs* to be for it to not utterly flop in the marketplace. If Apple can't figure out how to sell it there, then this product becomes the next Cube.
Exactly. I have $1400 in sunk costs on internal storage for which I'd probably be looking at putting on eSATA-on-TB or USB3. Sure, I can jerry-rig something near term, but the next turn of the lifecycle crank isn't going to be pretty looking.
"Military Grade" Aluminum? Really?
Try a COTS standard grade. IIRC, it is 7075, which has varied in price from $3 to $1 per pound. In China, there's less pollution regulation, so err towards the low end of that scale.
Sure, but there's still cables, mounting brackets...it won't be utterly free.
Which all depends on how pitifully small the stock SSD is that Apple offers. We both know that the upgrades won't be cheap.
A good breakdown ... and it is group (b) that gets hosed by the new MP...it is just history repeating itself from the dear prices for external Firewire ... and external SCSI drives before that.
-hh
Precisely. I think the design is as much to drive the price down as to do something new. They've dropped a lot of components and PC board space and all we've seen is the high end.
- If this is priced starting at >$3k as some say it will be DOA.
- If this is priced at $2.5k it will do OK, but probably worse than the present Mac Pro because it has less internal options.
- If this is priced at $1999 then it will do just fine.
I'll go on a limb and say it's possible that it will start even lower with bottom of the barrel GPU's, say $1799. That will really light a fire in sales.
Now, how much will it cost when you add your TB enclosures? Do you plan on running on USB 3 for everything, with all of the compromises that it entails, or will you be dropping $600 for a 2 bay enclosure or $900 for a 4 bay enclosure?
There are fat profits and then there are super fat profits. The workstation GPUs tend to have at least as high margins built into them as Apple makes on some their highest margin systems.
the thickness ( more so the military grade connotation) , seamless edges , and molded internals aren't going to be cheap to process. There grade/quality/etc. is most mainstream cases is geared to reducing costs. The vast majority of "well mainstream cases cost $80-100 so the Mac Pro cases can't be that much" estimates are off.
The Mac Pro design minimizes cables. And these are still sub $12 parts on a multi-thousand dollar box.
Small relative to what? pitfully small in capacity compared to the old 1TB HDD standard. However, that is completely Apples to Oranges to comparison.
The rMBP comes with 256GB SSD minimum. The Mac Pro is probably going to cost several hundred dollars more. You think Apple is going to put a SSD smaller than a rMBP minimum in a Mac Pro? I do not. Not very likely at all.
Far more likely your notion of sub $2,000 new Mac Pro price is wrong. Sure Apple could gimp the Mac Pro and shrink the SSD smaller than a laptop drive. But to what end? Too small and folks will question paying for the box. Hyperfast bandwidth isn't gong to make up for "ran out of space for my app's standard collection library".
Unless that group was dominantly dependent upon eSATA, there is little indicate they are going to get hosed. There are several factors in play. First, thunderbolt "sneaker net" drives are on the increase. ( they work fastest with field Macs with Thunderbolt much better than FW , USB 3.0 , or eSATA ). that is only going to get bigger as more of the deployed Mac user base transitions to new Macs .
Second, if "hosed" means will need external USB 3.0 / FW ports then the presumption that those won't be bought anyway (e.g., multifunction TB device that adds ports and does other things). And nevermind that FW daisy chains and USB expands ( folks who have a large stack of USB HDDs probably have a hub somewhere too ).
There are 6 TB ports worse case folks could throw one of those at a TB->FW dongle and still have twice as many left as on any other Mac.
If referring to external TB storage devices. That is primarily an issue with their design and not inherient to TB . Most TB storage focused devices are designed to bundle RAID cards and/or HDD/SDDs with the box. The prices are in part driving substantially higher by the bundling. Thunderbolt is there too ( adding infrastructure to break-out DPort if a legacy device gets plugged into the second port adds costs ) but that isn't the only factor in the higher prices.
There aren't any pure JBOD devices. And switching from the RAID/eSATA card bought to one bundled inside the TB devices isn't too much different that transition from old style SCSI to SAS. Sometimes technological inflection points means moving forward on new equipment.
I think it's going to do better than some forum members think. I just hope that I can get into the mid-level version for under $3K.
Several factors negate that which has nothing to do with folks dumping previous Mac Pros in mass ( they aren't, nor are they expected by Apple and most vendors to do that. )
1. Many of those 1,1-5,1 have already moved. Those that have moved become less likely to move again. No, the market will not dry up immediately but it will start shrinking and that shrink rate will accelerate over time.
2. Some of those 1,1-5,1 are going to die. A dead Mac Pro isn't going to get a new card. Actually a dead Mac Pro is likely to produce more used , but "newer" GPU cards that folks want to sell at a discount. This too will eat into his market. Same thing with folks dumping Mac Pros with card upgrades and going to WinPC system makers.
For the ones that don't die another substantive subset will be put on the "practice squad" or the bench. Relegated to file serving, back-up server, fall back spare machine , container for my old 10.4 image to run PPC code, etc. they aren't gong to be moving forward much on video card.
3. Newer cards are going to require drivers that only appear in newer OS. 1,1-1,2 are already left behind without a hack. 1,3 are probably going to be axed on next OS X iteration. [ 10.9 not leaving any mac behind is probably a temporary lull before they start reducing again. Apple needs more folks off 10.6. ]
Old Mac Pros with the different graphics model in their EFI are increasingly not going to be supported on newer cards. The gap between WinPC UEFI boxes and newer Macs is shrinking. However, that also means the gap between older Mac Pro and that new joined group is growing. Folks who buy into the delusion that having some empty PCI-e slots will always save you from future changes in direction of technology are going to be disappointed in a couple years.
4. This will probably increase the amount of driver optimization that Apple has Nvidia and AMD to just what embedded designs that Apple buys.
I see where you're going, but we do have to be cautious about quantifying real world costs. Don't forget that for all products that there's Fixed Costs and Variable Costs, and when one gets into specialty items with relatively small product volumes, the denominator for Fixed becomes small...for example, if there's $500K in 'Fixed' product development costs and you only expect to sell 2000 units, if each unit doesn't have at least a $250 markup versus its Variable manufacturing cost, you're going to be losing money.
Oh, there's no doubt that the entire Mac Pro case is worth more than $100...my point was simply that making that case ~5" shorter by removing the HDD bays/etc isn't going to reduce its manufacturing cost by all that much...call the effective 'value' of the case shrink to be probably around $100. In any case, I don't really think that the manufacturing cost difference between a ~4mm thick sheet and a ~10 gage sheet (thin) is really all that significant...and in some respects may actually have offered lower fixed manufacturing costs.
I'd like to be optimistic and believe that it will be a 256 instead of a 128...
...but I also know that Apple has chronically undersized the default OEM storage, as well as clobbers consumers on upgrades.
I agree, yet I nevertheless really do expect Apple to lobotimize the product somewhere to hit some magical "Starting At" mindshare price. If it isn't the SSD, then it will be with RAM slots or something else.
Sure, but a good chunk of what I was alluding to here is simply the small guy who today is sitting with, say, four 4TB HDDs in the internal bays ... they can use the software RAID0 & RAID1 within OS X right now and not even have to fork over for an eSATA card and enclosure...their capability only cost them four bare HDDs.
For this use case to roll out an incremental upgrade to a '13 Mac Pro and accomodate the change in its form factor, they'll at least need to get an eSATA RAID box (or two), TB-eSATA adaptor(s) & TB cables to repackage their existing (legacy) internal HDDs into...plus whatever they were using for a boot SSD very well might be not cost effective to transfer (eg, spend $700 for a PCIe card box to hold an 256GB OWC Accelsior PCIe Card worth $500).
True, technology marches on and there's always casualties .. and some transitions invariably are more costly than others... the pragmatic business question for these customers is "What's My Upside Gain?".
If there isn't much upside gain for the customer, then they're behavior will be to postpone purchasing a new Mac Pro for as long as they can...or switch to a Windows tower. It always comes back to the question of the specific use cases and the optimal solutions thereof with today's (vs tomorrow's) hardware solutions. My general observation here is that "Group B" looks like they've really been hung out to dry by Apple's choices.
-hh
You don't necessarily need TB for spinning disks unless you're running large RAID arrays in which case, yeah, it could cost you a bit. But for the guy just migrating a few HDs from one Mac Pro to the other, USB 3 is plenty fast, and there are nice solutions for under $100.
If you've got a bunch of SSDs you want to migrate, then Promise makes the J4 TB enclosure for $370.
If you bought an overpriced Accelsior
p) you're gonna be happier with the new Mac Pro's PCIe SSD anyway but if you insist on keeping it, you won't mind having to buy an OWC Helios enclosure for it for $349.People just need to be smart about how they migrate.