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.
NAB 2012 and the Mac Pro
- Thread starter 24Frames
- Start date
- Sort by reaction score
Right, infiniband 4x QDR links are fairly common among the super computers these days and have 32Gbit/s transfer rates. Thunderbolt is sitting at 10Gbit/s. So you could make a cluster out of Macs connected via thunderbolt, it would be about as fast as the slower DDR infiniband, but why? Cost to performance trade off this versus a Sandy Bridge E Mac Pro just wouldn't be worth it. Maybe if Apple started using the i7 extremes that have 6 cores and can support up to 64 GB of RAM (since you'll be dealing with distributed memory, you're going to find some applications that need large memory on a single node) it could kind of work, but I wonder how an iMac would do with a 130W TDP processor?
So much writing on the wall
Really if Smoke is being retooled for iMac and Macbook Pros, then I really think the writing is on the wall.
http://www.larryjordan.biz/app_bin/wordpress/
"They also said, in a fascinating eyebrow-lifter, that Smoke has been retooled to run on the most recent generation Apple iMac and MacBook Pro systems. Though a Thunderbolt storage system, or other high-speed storage, is highly encouraged, editors are no longer forced to run this system only on high-end MacPros."
Really if Smoke is being retooled for iMac and Macbook Pros, then I really think the writing is on the wall.
http://www.larryjordan.biz/app_bin/wordpress/
"They also said, in a fascinating eyebrow-lifter, that Smoke has been retooled to run on the most recent generation Apple iMac and MacBook Pro systems. Though a Thunderbolt storage system, or other high-speed storage, is highly encouraged, editors are no longer forced to run this system only on high-end MacPros."
Very interesting. Either they are just trying to broaden their target base, or they know something is up. I'm betting on the former.
It looks like Apple has a strong commitment to FCP as a pro app.
It's a bit harder to read into this Mac Pro stuff. Editors have had laptops for a long long time. I wouldn't say retooling Smoke for iMacs and Macbook Pros alone is worrying, but the general theme of everyone doing it is worrying.
I think this may be more of a "go to where the customer are" kind of situation. If their data says that most of their customers are using high end iMacs or MacBook Pros (or at least own one in addition to their Mac Pro) then the correct decision is to retool to make sure you can target enough potential customers to make money.
Before hand it may not have been technically feasible to do so but now these systems finally have the resources to actually run the software reliably.
In my opinion and if I were writing consumer (non enterprise) software my target would not be a Mac Pro but the MacBook Pro. With hooks to take advantage of the extra resources a Mac Pro offers if present. Obviously some types of software preclude this option but whenever possible I would target the lowest possible system that can run my application.
Distributing processing load over 8x or 16x PCI-e v2 (or even better v3) links is far more effective bandwidth and latency wise. Other news out of NAB:
"... We had just gotten our Maximus system and decided the only way to meet the deadline was to put it into production right then and there. We used it to run Adobe CS5.5 and Autodesk 3ds Max and completed the job with time to spare. No way could we have done that without Maximus technology." ...
....
“By combining a single NVIDIA Quadro and Tesla GPU with Resolve, customers will be able to enjoy incredibly high performance color correction features, such as the ability to process an uncompressed RGB HD resolution video with five layers of real time color correction and image blur at 24fps. With a single Quadro and four Tesla GPUs, colorists using Resolve are able to color correct more than 20 layers of HD video in real time." ... "
http://news.creativecow.net/story/868550
Thunderbolt can't do any of these at those levels of performance. There is certainly decreased sacrifices because have deployed into the field with TB.
While a new Mac Pro still limited to 4 PCI-e slots won't hold 5 cards it still could do 3 with some new baseline adjustments to power/cooling in the PCI-e cards' thermal zone.
Last edited:
The writing is on the wall for Autodesk to make more money with Smoke.
They have also decreased the price to $3,500 from $15,000. It is very simple. They want to sell more copies at a higher volume to make more money. That's it.
A user base of 2,000 @ 15K/copy is $30,000,000
A user base of 8,600 @ 3.5K/copy is $30,100,000
The second is both more money and a more stable user base. (i.e., if 300 users bolted, went bankrupt, or otherwise stopped buying that has less of a risky impact on the second case than the first. The financial stability of your customers is largely outside of a software vendors control. )
There is no pragmatic way of significantly expanding the users base by restricting the software to a single Mac model. The Mac market is so relatively small that it is a bad idea for a software vendor to restrict their software to a small subset of that market. Sure not being tied to the Mac Pro reduces risk if it were to be canceled, but the small user base is also a large risk.
Why is it worrying? Obviously these companies can see which way things are going and are happy to develop new tools/software to accommodate.
All this mac pro whining amounts to a bunch of cry-babies having to use the same computer as Uncle Bob in the future. Their "pro" status will be in the mire if they are seen with *gasps* an iMac.
Get over it.
Apple has a clearly stated policy from several years back that they don't manage product introductions to 3rd party shows. It was always highly dubious that Apple would announce something at NAB, because they don't announce at shows. The FCPX demo last year was a not particularly well thought out exception case... not the rule.
The "goal" posts aren't moving. What is happening is people keep playing different games with their predictions. If it isn't part availability or an Apple show it is an entirely different game with different rules with different goal posts (probably ad view generation agenda). Even the regularly scheduled Apple shows are highly doubtful for a specific Mac product.
The "new Mac Pro " coming back in June/July '11 was loopy because there were no substitutable parts. If Intel had hit their dates in Q3/Q4 that would have made sense. When they blew past them expecting something different from Apple was dubious part.
The parts slide has impacted all vendors in the market for leveraging the parts.
If Intel , AMD, and Nvidia all colluded to put their parts on top of NAB then perhaps Apple's announcement would coincide with the show. That didn't happen this year. Nor is it likely to happen most years for the Mac Pro.
It's worrying for two entirely different reasons. Either:
1) They've already gotten hints from Apple that the Mac Pro is going away and they're retooling.
2) They feel customers are abandoning/no longer require the Mac Pro, which gives Apple more reason to discontinue the Mac Pro. Apple can point to these products as reasons the Mac Pro is no longer needed (which is dubious logic, IMO, but it's what they could do.)
Neither are for sure, but both possibilities.
As far as "whining", iMacs just aren't that powerful compared to the tools we normally use. It's not a status thing, it's a "getting our work done" thing. When an iMac ships with a pro desktop GPU, I will be happy to consider switching. As it stands, iMacs still ship with mobile GPUs running at half the speed.
People need to stop obsessing about the exact day. Everything points to something happening, the exact day is just a minor detail. Whether something happens today, Thursday, next week, or next month doesn't matter. It's interesting to speculate, but if you're going to freak out over Apple not releasing the Mac Pro in a certain week, it might be time to take a break.
Last edited:
The software vendors are in the software business. They are likely to follow the same trends since those trends are not going to be application or vendor specific.
If laptops and other Macs now have high speed, low latency storage and substantially faster GPU (equal in performance of what the software was targeted at 3-4 years ago) then why wouldn't they all expand onto those platforms? The more curious questions is why want they want to restrict the software to few models?
The notion of heavily silo-ed expertise and tool usage is also rather curious. Few of the companies/entities funding these hardware/software purchases typically want to overdose on that.
I don't know if the idea of people running pro software on consumer machines is new. The Powerbook G3 and G4 were both marketed heavily to FCP editors. I knew a lot of Photoshop users on iMac G3s. Yet the reasoning for an existence of a pro tower never changed. There were always people who needed more power than the basics those machines provided, much as it still is today. So in that respect, device makers targeting those users isn't surprising. Outsiders in this forum act like the Mac Pro is a dead device because one can simply install, run, and edit with FCPX on an iMac. But as it always was, FCP always ran on consumer hardware. And if you're a power user, the experience you need is usually only found on a Mac Pro.
But Apple these days doesn't seem to really care about the technical details. If the industry feels that Thunderbolt is the path forward, I think Apple would kill the Mac Pro and go along with, even if Thunderbolt is slow (for the sorts of things Mac Pro users require.)
Again why is it worrying? It's not important at all if Apple doesn't make a workstation. Plenty of other places you can buy one. Mac OSX isn't an excuse anymore when windows 7 can run everything you need.
I could certainly understand if you had no alternative, but apart from being way too married to an Operating System you have plenty of alternatives and no reason at all for all the complaining.
Unless your workflow/software requires OS X, like mine. Then I'm stuck. As a Mac/iOS developer, my tools are made by Apple and therefore not available for Windows/Linux.
So no, Windows 7 can't run everything I need. And yes, I have no alternative (besides Hackintosh, which I've talked about, or pitching a fit.)
Plus tools are not free. Even if you have equivalents on Windows, you're paying a pretty heavy price to crossgrade/switch. And honestly, even if I was in the position of having tools available on Windows, I use the Mac because I like the Mac better. Moving to Windows would still be a step back, even if I could still do my work.
Yep. But those are the cards I've been dealt. I suppose I could switch to Android/Windows development, I'm sure that would make me a lot of money...
...I'm sorry, I can't stop laughing now.
Ironic that the people who make the software so tightly integrated with the hardware are going to be hamstrung by the hardware.
Assuming the Mac Pro is EoL'd, what are you considering? Hackintosh or high-end iMac?
Not sure. I really enjoy the number of cores in my Mac Pro (XCode is an app that can double it's speed with double the cores), and I do graphics/CUDA/OpenCL programming, so the GPU is really valuable to me...
I could probably make do with 4 or 6 cores. It would be slower but I could deal with it. The GPU is a bit more tricky.
I'm mostly keeping my eye on what happens with the next Macbook Pros. If they're high end enough, I could go with one of those and a 27" TB display. Wouldn't be as good as a Mac Pro, but better than nothing. Otherwise, a max'd out iMac and Macbook or some sort is probably in the cards. Again, better than nothing.
I've also been considering building a PC as a secondary machine, I wouldn't be able to do most my work on there, but at least I could simulate GPU stuff in order to target next year's Macs.
Thunderbolt is two 10Gbit/s bi-directional with aggregate of 40Gbit/s. 4X QDR is 32 Gbit/s aggregate.
Agree, Thunderbolt is not as fast as CPU to RAM which is around 25GB/sec (about 200 Gbits/sec). But it is viable to daisy chain up to 6 devices -- there are short distance limits too.
But whatever comes out of this, if Apple think they can continue producing MacPro's with the current rate of updates (every 2 years) that's a dead end road.
How arrogant and rude. And ridiculous. Not to mention highly ironic.
Ironic, because:
Your whining ABOUT mac pro whining boils down to having nothing better to do than stroll around an anonymous interweb for a while and incorrectly assert that the frustration expressed by hundreds of people (none of whom you know) is completely fabricated and motivated purely out of some superiority complex. Talk about the pot calling the kettle black... Apparently, without ever having exchanged a single word, you not only think you know me and my computing needs, you also know my motivations, and have fired off a couple of general and petulant insults because... why, exactly? Of you and me, just which of us needs to "get over it?"
Seriously- if someone says they need a Mac Pro, and that they will be in a tough spot if the line dies out without a suitable replacement, why on earth does that bother you? Why on earth would you even care? Simply because YOU (clearly) don't need one doesn't mean somebody else doesn't. Why on earth do you feel the need to argue against and put down someone else's situation? It's not like the existence and availability of a mac pro is somehow going to hurt you, is it? Apple's other products aren't going away. So what possible dog do you have in this fight?
And for the billionth time on these forums, *I* need a Mac Pro (or suitable equivalent). It's got nothing to do with being labeled a "pro" or not- it's got everything to do with needing to use my software tools efficiently alongside the rest of my hardware tools which interface with my computer (hardware which neither an iMac nor a Mac Mini can connect with, BTW).
I NEED (not prefer, but NEED):
- multiple PCI slots
- to be able to run my main software packages (at least three of which are OSX-exclusive, and the rest are obviously all OSX versions)
- many computing threads for real-time audio processing
- expandable/replaceable RAM
- swappable and replaceable internal and external storage
- a near-silent running machine
- quality cooling
- optical drives
- tons of I/O ports, peripheral options, and expandability
- high I/O bandwidth
- a non-reflective video monitor
- a (mostly) user-serviceable machine
Hey- if that happens to match up with a machine that Uncle Bob is using, no problem. Give me what he's got. They can even call it the Mac Hobbiest for all I care. The name doesn't have anything to do with why it is useful to me. But it better have the capabilities I need.
I've never in my life delivered a project or demoed my music to someone (client or otherwise) and said "hey, this was made on a very important and exclusive professionals-only computer." Why would they care? Why would I care, for that matter? That's just stupid.
My last thought:
If there's a "look at me" vanity segment of Apple's customer base, I'd suggest it isn't from within the power user community. Pros (and semi-pros and power hobbyists, for what it's worth) use Apple products as tools. They know what they need. Their computing purpose is pure utility and ROI, and they invest with a plan and purpose in mind. However, it's pretty obvious that a sizable percentage of "iConsumers", some even on impulse, buy shiny nicely-packaged toys with the primary purposes of being noticed, having a current/trendy status symbol, and owning the latest conversation piece. Nearly all of said iConsumers don't actually *need* the vast majority of the things their iDevices of choice provide.
I think I've made my point(s). Sorry for the long post.
Smoke to run on iMac's and Macbook Pro's? So what? Always could do it. Just slower. So they made it lighter? Why would they not do that anyway regardless of HW being used? It means nothing. Look at what mobile chips can do now vs 2 years ago. Of course it is viable to run this stuff on quad chips that can hit over 10,000 on geekbench. Only the most recent Mac Pro's and iMac's get over that. Just because mobile is getting faster does not mean desktops will vanish. Unless we just want to tread water on performance for a few years and be hamstrung to thermal limits at all times.