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.
People Who Complain About the New Mac Pro Misunderstand It
- Thread starter Radiating
- Start date
- Sort by reaction score
The issue isn't data. It is methodology. If the methodology is jacked up the data isn't even the primary issue.
But common fracking sense. Lower priced boxes tend to sell in higher quantites than higher priced ones. The claim here is that Mac Pro users in mass skipped over the far more affordable $2,000-3,500 boxes to go to more expensive ones. Frankly some of the top moaning and groaning is about costs. "folks won't use TB external boxes ... they cost more". but when comes to buying the Mac Pro all of a sudden a $1,000 more doesn't matter to the broad spectrum market? Yeah, sure I have a bridge to sell you. It is bright red/orange with scenic views.
The thing is, the single socket Mac Pro has always been a poor value relative to competition.
The dual socket has always been a solid value.
To me, the "common sense" would be to buy a machine that's a good value.
I gave an anecdotal statement that of the 100 or so Mac Pro's I have deployed over the years, or among the motion graphic and 3d crowd that I interact with, I can't remember anyone going with a single socket machine.
Wrapping your anecdote in an authoritative tone of certainty doesn't make it any less anecdotal.
Like that?
That would be why I don't pay as much attention to the "up to" specs. It's more interesting to know what falls into the realm of what I would actually purchase.
I think a lot of the moaning isn't about the money per se than it is that it is an unnecessary expense compared to retaining even the mediocre slot configuration of previous models. (not even considering competition...). At least it is for me.
To get any sort of expandability one has to shell out additional money, create additional points of failure, have additional boxes sitting on their desk, all for the privilege of reduced interface performance.
Or they could have made a workstation that is, you know, competitive with their competition from a performance and I/O standpoint.
They've chosen to make an appliance, and that's cool, but long-time high-end Mac Pro users are just disappointed.
What are the chances that with the pre-showing of the new Mac Pro design, Apple is watching how sales of the still-available tower Mac Pro (except in the UK) develop until the new one is available? If there is a significant sales spike indicating continuing tower interest, Apple could update the tower Mac Pro as well and offer both to the consumer. They could also fix the fan issue so as to resume UK sales at the same time.
Apple never actually said the tower Mac Pro was discontinued, did they?
Apple never actually said the tower Mac Pro was discontinued, did they?
Correct, so if someone doesn't need two CPU packages they aren't gong to buy them. We are far past the stage where need to buy two packages to get over mid-digits core count. We are also well past the stage were need to buy two packages to get to relatively affordable 64GB of RAM. Therefore, not much need.
A photographer who needs a Mac Pro for his studio fits the profile. Developer? Fits the profile. Smaller business that needs "more than an iMac' and/or 'longer probable lifespan than an iMac' ? Fits the profile.
There are more than several broadly populated groups where single is good enough. There are more than several narrowly populated groups were single is not enough.
The rule of thumb "pros need dual package" worked quite well 5-8 years ago. Technology has changed though. For a big chunk of that period 'dual packages' was a significant distinction of core count and memory needed for above average workload. With memory density increases and core count increases it is not.
Dropping dual means not covering all of the previous set. But if have most of the previous set it isn't a big issue.
If far more than 50% of the folks who bought Mac Pros from 2009-2011 were upper end dual folks then Apple would have dumped the single option. They didn't.
The memory capacity is expandable over time. The storage capacity is over time also.
Frankly, expanding to additional devices adds additional points of failure whether internal or external. That actually particularly significant. If extremely sloppy about cables maybe a marginal difference, there is not requirement to sloppy.
They don't have to be on the desk any more than the current Mac Pro has to be on the desk.
For PCI-e cards.... relative to the current internal SATA controller there isn't. And folks survived with it.
Yes for cards that push x4 to the limit and x8 and above there are tradeoffs. But again you'd have to look at the broad market demographics.
If their competition was much more profitable and growing faster that actually might be a bigger issue. Find a growing group of people and sell them something they are happy with and you make a profit selling it. That is pretty much the objective. Specifically, beating Dell or HP or Lenovo up isn't necessarily the primary objective.
" It just works " has always been where the Mac was primarily going. And yes I do realize if there isn't "enough" it won't just work, but the question is just how much is "enough" to cover most folks. Not how much is "enough" cover everyone.
I think a decent number of folks who used to be "high end" Mac Pro users aren't at the high end anymore. Their workload has started to platuea and the technology has jumped in front. They are on a different curve now.
Just like on the HP/Dell/etc side those folks are moving from 800 class machines down to 600 and 400 class machines because they are good enough.
Those folks splitting off is a major driver here not some "I'm going to kick the top end folks in the shin for giggles" motivation. Apple is following them and the growing trend to externally based bulk storage.
Exactly correct: the pragmatic perspective seeks 'Best Value'.
Yes, which is also simply the "Best Value" metric becoming evident again. For example, it classically has been that an additional internal drive (eg, storage capability) is both less costly and more pragmatically reliable than an external, which manifests the internal as the better value of the two options. This trend continues with the new Tube, due to the "Thunderbolt Tax" which is applied to any so-attached peripheral.
-hh
Watching sales to see what tell the factory each week to make just enough to sell? About 100% That is what they do for every single product they have. That's is why they have some of the leanest inventory of any PC/tech company in any sector.
Watching sales as some "product market focus" testing exercise. About 0.00009% There is not alternative E5 Xeon v2 tower design hiding in the closet that was developed as plan b. If they had been doing plan b's like that they could have shipped in June 2012. They didn't. In fact it was same GPUs.
The current design is already in the non viable stage ( withdrawal from EU markets ). Intel is about to pull the processors used in the device from the retail market this Fall. ( apple can still get replacements for a year or so ).
I think they need to closely watch because TB v2 chips are not ready to go right now at all. Maybe by Fall but not ready. Likewise Intel may/may not slow roll the E5 1600 v2 out the door.
Pretty much underestimated how long and protracted the upgrade process is. If Apple wanted to do a tower now it probably 12 months till it showed up. That's relatively awkward because Xeon E5 would be creeping up on another possible transition move to a new socket socket and microarchitecture. If start now it makes more sense to target Xeon E5 v3. But a very real chance E5 v3 will could slide into 2015.
That is simply because they didn't start selling the new one or even hint that they will be taking any orders soon. They aren't because it is highly likely months away. The SNAFU with the iMac last Fall was that they stopped it before it was ready.
The sneak peak serves three major purposes.
1. Effectively is it like the XServe cancellation where Apple gave folks about 3 months to buy them if needed them and wrap up in flight purchase orders. By the time Apple stops selling the current one new no one will have any possible excuse as to how they didn't know a change was coming.
Effectively the dual processor variant of the current Mac Pro has been cancelled. If need one, get one. Otherwise going to be in used/refurb line.
2. To software developers (appropriate at a WWDC ). "No we haven't been joking about OpenCL for the last 3 years". The whole mac line up is going to leverage this. At the top end even more. There is a developer theme that this box hit so it is in the keynote.
3. It is a huge surprise for many. By the time this comes out and lots of folks can kick the tires and start to rationally figure out if this works for them or not. It might also allow Apple to bring in more software developers for test drives without having to resort to stuff the Mac Pro into a filing cabinet box.
Some larger and/or beta testing clients will get pre-production units to work out some bugs.That is far more in Intel's hands than Apples. For whatever reason Intel screwed up the 3500 - 3600 transition.
For machines named Mac Pro the "single socket" version has pragmatically really only been from 2009 forward. Before that you HAD to have multiple packages to crack 2 cores.
Frankly that screw up probably didn't help with the internal debate over continuing either. It wasn't until early 2012 that Intel fixed that issue. Sure it has been a year but Intel missed a year also.
Which is crappy methodology for inference about the whole Mac Pro market. You have admittedly biased sampling. There is no way it accurately reflects the market. A very narrow subset of the market, perhaps. The overall market, no.
Sampling prior to 2009 is goofy because the single processor Mac Pro pragmatically didn't exist until 2009.
Your methodology being deeply flawed is not an anecdote. It is.
You haven't seen a single CPU Mac Pro because haven't been looking for one.
Yes, there has been a contemporary plateauing of hardware needs in a lot of IT use cases, and this has resulted in some customers being able to shift down from bleeding-edge hardware. But by the same token, there's still a contingent who will pay for more...its just a question of if these are adequate to support a viable business case to deliver the goods.
Similarly, when viewing these questions, we also need to be cautious to not permit "core count" to become the new 'MHz Myth' when trying to equate discrete features to the capabilities that they represent.
And you're halfway. The second half is to examine what it is that these groups need which is "more than an iMac" and then determine if the Tube is able to deliver, or if it falls short...especially from a "Best Value" paradigm.
Oh, but if that was the only variable.
For example, one's "Starting At..." advertising copy is also an important factor in attracting customers, as well as fostering an upsell environment...just as we also have "Up To..." performance claims too.
Neither of which is a new capability. The only real question comes down to how much does it cost to afford that capability...and (of course) if that is a better value or worse value in comparison to its own history (and competition).
I think I understand what you're saying, but from a technical perspective, an external implimentation always has more failure points than an internal. For example, consider a simple hard drive and just count connection junctions: there's currently the connection point inside a bay ... which we can assume is equal to the HDD connection inside of its external case ... signal and power. But the internal is now hooked up, and from the perspective of the external, the signal has two more cable connections (case to cable, cable to PC), plus we also need to sort out where its power is coming from. In this very oversimplified example, the external has 3x more failure points.
Fair enough, but the customer pushback is because they've perceived that Apple has functionally reduced the number of product options that they had to pick from. In addition, there's also a psychological impact in the product line's upsell potential as impacted by top-of-the-line "asperational" products...IIRC, it was probably the Harvard Business Review that discussed why the Corvette in the context of the Firebird/Camaro at GM.
Except the ones who are getting kicked in the shins aren't the top of the food chain, but the smaller enterprises who lack fast external storage facilities.
Apple's business case is seeking ...legitimately... to minimize their costs and this desktop customer certainly isn't a strong growth segment with which to amortize fixed development & production expenses across. As such, one of the most effective ways to get these fixed costs down is by having fewer fixed costs - - and a product streamlining to go from two motherboard designs (dual CPU + single CPU) to but one motherboard design does accomplish that business objective - - even if it results in a modestly reduced sales volume. Add to that how the forthcoming CPUs reportedly are going to be available with up to 12-cores, the approach would appear to be justified - - and it very well may likely be just so ... but only for that one particular metric.
-hh
LOL.
Dude, methodology? Relax.
I was making a statement about my experience. If you want to take that as some kind of prognostication about what I think of the entirety of the market, go for it.
If i say "The sky is blue."
You don't have to scream "YOUR METHODOLOGY FOR SAMPLING THE POSSIBLE COLORS OF THE SKY DID NOT INCLUDE ALL POINTS ON EARTH THEREFORE YOUR ANECDOTE IS WRONG"
Since Apple went Intel, cheap MPs don't sell any longer .
Not after 2008 anyways .
That's because the performance/price ratio is what matters, not the price tag .
A quad Intel MP is utter crap, always has been .
Low clock speeds are useless for everyone, low core count for quite a few .
The best bang/buck MP is the 3.33 Hex, which is a joke .
That thing is hardly faster my entry level MP 3.1 2.8 Octo, if I take my occassional 3D renders into consideration . Ram is cheaper, great, but GPU support is as crappy as it always was .
The new MP will be price tag + 1000-1500 for basic TB peripherals, and only if we are lucky will the price/performance ratio be a bit up (MP only, sans the TB stuff) .
That's a bridge right there, what Apple is going to sell .
You're going to get nitpicked to the point of insanity if you keep debating with him, that's why I stopped. He can never see your point and accept it, just pull every paragraph apart sentence by sentence for no reason (even when he isn't disagreeing with your point). He's made up his mind that TB is every bit as good as PCIe and if you don't like it than you're just not a true "professional," as professionals have an infinite amount of money, don't care about internal storage, and don't mind a "squad" (his words) of little boxes full of thunderbolt droppings around their desk. Classic "no true Scottsman" argument. If you don't like the New Mac Pro... well it just wasn't built for you, it was built for some mythical ideal of a professional he just came up with for the sake of his argument.
Apple's decided to make an iTube club for "professionals" with more money than brains who don't mind continually sacrificing for Apple's little pet features. My bet is that one of 2 things will happen: Either Apple will sell it at a loss and just flood the market with these things (which would put a huge boost into the Thunderbolt market) or, more likely, it will fail miserably while fading into the obscurity of the 20th century mac.
Last edited:
What...and what? 12 threads at 3.46GHz vs. 8 threads at 2.8GHz? 2 Threads at 3.6GHz vs. 2 at 2.8GHz. It is a crap ton faster in real world regardless of what the geekbench says. Oh wait...
8-core 2.8 = 7685
Hex = 13890
You'll need a 3rd processor to compete. At least in 32-bit.
And the GPU support? Same as any GPU support. It is great now. Where have you been?
5770, 5870, 6950, GTX 680, GTX 570, GTX 470, GTX 670. What else do you need? That's most of the cards you can buy within the power envelope of the PSU.
You sound angry about it. What SW are you doing renders in? Sounds like it scales terribly.
I agree that is probably the case for semi-pro and home users. They are naturally more diligent about their computer budget. I think most businesses buy the dual cores. Mac's in businesses tend to be held on to longer than PCs. So more cores is kind of "future proofing" Maybe it is the build quality too? That and in the last few years laptops got more acceptable as desktop replacements. The users that are still using Pros are the ones who specifically need (or think they need) more cores and ram.
Sorry, got a little carried away there .
The 3.33 Hex is definitely faster than my 2008 2.8 Octo in all applications, but not crazy faster .
The current quads are faster at comparable clock speeds, in non-multithreading applications .
Also, depending on the software you use, CPU hyperthreading doesn't equal physical cores .
I use Cinema 4D, and here the performance increase is 40-50% tops, depending on who you ask .
That makes the current MP Hex roughly equivalent to 8-9 physical cores, the Quad 6 cores .
Anyways, my point is, the MP 3.1/ 2.8 Octo was an entry-mid level MacPro, quite affordable and fairly close to the top model's performance .
It was, and still is, capable of running any available apps without much sacrifice ; clock speed, core count, all there .
2009 to now, different story .
There is no MP anymore that does it all, in a comparable price segment .
It's amazing how people will try to excuse everything and anything that Apple does or doesn't do. Let's face it: for professional grade applications when dealing with limited funds both the Windows and Mac side of the industry is becoming more and more difficult. Everything is fine for your standard home and business hardware and applications. In fact, it's great for your average consumers.
For anyone whose life is a little more complex things are becoming more difficult and expensive. I am a business person who works for a large company and my iphone supports the Exchange server on my iPhone more or less ok. For my personal business things are trickier and not well supported compared to the old Blackberry email world.
Then I have a passion for photography and music and I do some stuff in those areas expecting professional results. Finding a good monitor these days or a laptop with a good high resolution screen (and I don't mean "retina") is a problem and has become very expensive. Sound card connectivity is a nightmare.
100% compatible hardware components for these applications are also hard to come by now. Even software has become an expensive nuisance unlike ever before with for instance Adobe forcing on us their stupid subscription model that I will not support.
That's why I'm stuck with old computers and software at this point. I don't see a clear upgrade path, neither with Windows nor with Apple. Even worse if every few months some marketing geniuses and their supporters (see above) try to declare what we just bought obsolete.
I understand that Apple et al are trying to skim off the consumer market as much as possible. That's all fine. But pros are completely left out - at least those that can't go and spend thousands upon thousands on hardware and software any time they need to upgrade just a few items.
For anyone whose life is a little more complex things are becoming more difficult and expensive. I am a business person who works for a large company and my iphone supports the Exchange server on my iPhone more or less ok. For my personal business things are trickier and not well supported compared to the old Blackberry email world.
Then I have a passion for photography and music and I do some stuff in those areas expecting professional results. Finding a good monitor these days or a laptop with a good high resolution screen (and I don't mean "retina") is a problem and has become very expensive. Sound card connectivity is a nightmare.
100% compatible hardware components for these applications are also hard to come by now. Even software has become an expensive nuisance unlike ever before with for instance Adobe forcing on us their stupid subscription model that I will not support.
That's why I'm stuck with old computers and software at this point. I don't see a clear upgrade path, neither with Windows nor with Apple. Even worse if every few months some marketing geniuses and their supporters (see above) try to declare what we just bought obsolete.
I understand that Apple et al are trying to skim off the consumer market as much as possible. That's all fine. But pros are completely left out - at least those that can't go and spend thousands upon thousands on hardware and software any time they need to upgrade just a few items.
My guess is the 3,1 2008 Mac Pro was one of the popular model as most users bought this as best bang for it's price and specs. I am also using a 2008 Mac Pro until now and you're right that it's still dependable up to now.
Like this kind of stuff had not been highlighted years ago....
Non-linear improvements for every app in the benchmark suite. ( 4 to 8 same architecture )
http://www.barefeats.com/nehal08.html
Hex? in 64-bits the 2009 4 core beats the entry 8 core line up from 2008.
http://www.barefeats.com/nehal03.html
The 2008 and earlier front side shared bus is a bottleneck once start to do anything substantive that requires pushes the memory I/O bandwidth limits.
(e.g., embarrassingly parallel floating point operations on all cores concurrently).
32-bits and limited to the also crippling small register set and the 2008 have more enough bandwidth throttling constraints to look good. So yeah if primarily interest in running 2006 OS X 32 bit intel code the MP 2008 models are a better choice. If actually trying to track software from the present decade it isn't.
System $/performance metrics can easily be swayed is skew a higher percentage of the systems price to be based on RAM costs and then trade "last years, high volume" affordable right now density versus bleeding edge densities. Over time over the whole Mac Pro (on Intel) evolution the single CPU offerings have been expanding.
Typically highly correlated with software deeply grounded in the past.
aside: forgot about the 2008 dual capable but single socket 2008. So slight more coverage of offering over 6-8 years , but given only one of three and lower gap between entry marks in single/dual the duals did sell higher number than later specific designed singles did.
Last edited:
You had me until you said sound cards. There are pretty decent class compliant sound cards that are darn cheap. Then there are the USB and Firewire equipped mixers. Both Dj and console style. And at the high end we get UAD and Apogee boxes with open configuration architecture that allows different output modules to be fitted.
But then I have had a MOTU 828MKII since they came out. Still works pretty much perfectly from any Mac.
And then I gotta say there are some really nice monitors from boring old Dell and HP. May not fit the "style" of a mac, but they have really good color gamut and accuracy.
Audio is one of the apps that still have solid support for the Mac. Throughout all market segments there are a wide variety of interfaces. The Apogee stuff is quite nice. Our composers and musicians use it quite a bit. We use a variety of gear from RME, Avid and other smaller high end boutique style manufacturers. At the high end the pre amps and converters are outboard and have been for years. All you need is good digital I/O. On the lower end there has never been a better time to get good gear at a very reasonable price.
Its kind of funny that I forgot Avid. After all my work has a gigantic Avid iNews/Newscutter install. But on the audio side we use SSL digital consoles to get the signal into the digital domain.
Frankly Avid/Digidesign has a history of mediocre audio quality.
Thats what got Black Lion Audio in to business. Modding Pro Tools interfaces to sound better.
A lot of us would like to forget about Avid...
The fact is that if you need to do large track counts (96-128) and use Pro Tools you're pigeoned into PT interfaces and cards. I haven't looked at PT 11 too much as we're still on PT 9 Native and got off of HD/TDM about 6 years ago after working on several large track count (96 or so) TDM rigs. Native HD can do MADI but we have two Hammerfall cards in our large rigs. The weak link in the input chain are the pre amps and converters and I've been doing outboard pre amps and converters since we went from tape to DAW completely about 10-15 years or so ago. We only need something to take a clean digital signal.We went to Nuendo on a PC for our biggest rig because at the time the Mac version was pretty flakey and the post houses with which we work use Nuendo. We saved 40-50k in interfaces alone and depending on who's ears are in play some say the engine sounds more transparent. We still have a wide variety of Pro Tools rigs but most are in the 24 track and under capability. When we do something for broadcast we send it out with our people to a big room for sound to picture and mix the music only projects in house at HQ with an old school Neve/Pro Tools hybrid system. I think they still have an A-810 and a couple of A-827s up there, yep, we're old...