Next MacPro: Predictions...

[M Most]

Aha, I copy you now. And this is how the suggestion of using LP as the interconnect has come into the equation?

Yes. In my original post, I said "perhaps using something like LightPeak as the interconnect." That in turn launched a discussion about LightPeak - not my original intent for this thread, but I have no problem with it :thumbup1:

 

[ericyoung]

The lament was, that he would like to have the capability right now, which is indeed available today, if needed. As far as future proof, not sure what you mean. Why do you think, that 10GB fiber has no future road map? Light peak doesn't have monopoly on that.
Now, if you mean lower price of interconnection, based on economies of scale, than that's different story altogether...

I meant that Lightpeak should have a relatively long service life as an interface standard, not that it would necessarily supplant 10Gb fibre.

And yes, I meant it would be cheap and ubiquitous.

 

[Dustin Cross]

Lightpeak is 10Gb/s

PCIe v2 is 4Gb/s per lane

So a 16 lane PCIe v2 slot can do 64Gb/s. A Mac pro currently has 40 PCIe v2 lanes. That is 160Gb/s. We just can't get enough cards in the thing to use anywhere near that much bandwidth.

Lightpeak needs to get a lot faster if it is going to be the only interconnect for new systems. And the 100Gb/s they are planing by 2020 is not enough.

High end systems still need PCIe and dedicated memory and and and, so turning the Mac Pro into an appliance with a couple Lightpeak connectors on back isn't going to happen.

The spec for PCIe v3 is done, is twice as fast as PCIe v2, and devices should start coming out next year.


Dusty

 

[M Most]

High end systems still need PCIe and dedicated memory and and and, so turning the Mac Pro into an appliance with a couple Lightpeak connectors on back isn't going to happen.

I think the bigger question is whether Apple is still going to manufacture what you're describing as "high end systems." Or if they see a need to. They were, at one time, pushing clustering as their ultimate answer to more computing power, so it's not clear to me that they see an all-inclusive single box as a solution, either. Hopefully the answer to all of this will become clear in the next few months or so.

Of course, for many of us who need REAL power and stability, there's already Linux... :biggrin5:

 

[Peter Mosiman]

Of course, for many of us who need REAL power and stability, there's already Linux... :biggrin5:

Is Linux what you guys run at the office? I've never really looked into Linux, but with the forthcoming OSX Lion, I may start having to...

 

[M Most]

Is Linux what you guys run at the office? I've never really looked into Linux, but with the forthcoming OSX Lion, I may start having to...

Our entire facility is basically Linux based, although we do have a rather healthy number of Macs around running things like Avid, Final Cut, Pro Tools, Smoke, and the like. But the infrastructure is all Linux, all of the conforming and grading work is done in Baselight (which is Linux based), and all of the internal software (on which much of what we do is based) is developed and deployed on Linux as well. And much of the time when you see someone on a Mac, they're doing something using command line in a Terminal window (including me). The only Windows presence is desktop machines for corporate stuff. You'll find Linux infrastructures in a lot of large Los Angeles facilities, especially those running either Autodesk or Filmlight systems, and particularly in large visual effects facilities. For efficiency, stability, and reliability, it's the best there is right now.

 

[Thomas Mathai]

Well if you look at a timeline of Apple products, the Xserve came out in 2002, when Apple was still Apple Computers, not Apple Inc. I think back then they cared about the enterprise market, not so much now.

Xserve was also I think the first product to have the aluminum case, a design philosophy that carried over to the other products until the iPhone, after which I think everything started having the black and aluminum style.

The question is does Apple want to make computers for a professional market. I personally think yes. At least a general professional market, nothing as specific as enterprise servers and storage.

To me, the iMac design is simple and closed, just like iPhone, iPad and laptops. Adding modules isn't part of that design, and I don't see them doing it even as a pro machine.

I've never seen them attempting to use a one size fits all form factor across their product line since they left beige behind. I have seen design elements of a new product show up in an different product line. I think when they see themes that work, they try it on other products.

The only thing I'm sure the next Mac Pro would have is more rounded corners. Maybe they could take elements of the mini, make it more squarish.

Guess we'll know when they unveil them.

 

[M Most]

I just wanted to thank everyone for their thoughts on this thread. One of the main reasons I put it here on RedUser is because whenever I talk to Jim, or sit through one of his presentations, I find myself inspired to think outside of the box, to throw away the rule book on a lot of things that I often take for granted. It's at those times when the most inspired and creative ideas usually happen, whether it's in regard to one's work or simply personal interests. I found myself in one of those moments the other day after having lunch with a friend who is deeply involved with 3D conversion work, but who's also a visual effects veteran (like me) and with whom I have a lot of conversations about technology in general.

Clearly, speculation about Apple and their future plans is, at its heart, somewhat pointless, because Apple is a very secretive company when it comes to future products, and also because nothing we say is likely to influence them, especially when most of their roadmap for the next few years is probably well established already. But it's fun and sometimes inspiring to play the "what would I do if I were them and had no limitations" game. So I thank everyone who contributed. Interesting points made all around.

 

[Jody Neckles]

Lightpeak is 10Gb/s

PCIe v2 is 4Gb/s per lane

Dusty

Apparently Light Peak is scalable to 100Gb/s. If and when this will happen is anyones guess.

 

[Dustin Cross]

They plan on 100Gb/s by 2020. Not soon enough.


Dusty

 

[jake blackstone]

grading work is done in Baselight (which is Linux based)

For sake of complete information, Baselight runs on both Linux and OSX.
Also, Filmmaster is well respected and widely used high end grading system, that runs only on Windows. It is used and well integrated in many top end post shops...

 

[M Most]

For sake of complete information, Baselight runs on both Linux and OSX.

I don't know of any shop that's using the OS X software for full grading and/or finishing, it's primarily out there as an on-set tool and on occasion a learning tool (as well as being the basis for the Kompressor module, but that's another story). Do you? As far as I know, it doesn't support any hardware surfaces, which limits its usefulness as a production tool even if one did want to use it that way.

Also, I don't know why you're bringing up Film Master. The question I was replying to had to do specifically with Linux. My answer wasn't a review of grading systems, it was a comment on the significant penetration of Linux in our industry.

 

[Marc Wielage]

There's been a lot of speculation and scuttlebutt about the future of the Mac Pro over on the LA Final Cut Pro User Group. The insiders I know tell me that there will be at least one more Mac Pro next summer, a 16-core Intel (upgrading the current 12-core), and probably one beyond that for 2012. After that... who knows?

Apple does seem to be becoming more and more a consumer electronics company, concentrating on phones and devices like the iPad, and getting away from their original philosophy as a computer and operating system manufacturer. There were many who were afraid that Apple was going to move away from Final Cut Pro and their other high-end applications, but apparently they're going to stick with them for now.

No question, these are very small markets compared to the other businesses they're in right now. I bet they make more profit with one month of iPhone sales than they've made in 12 years of selling Final Cut Pro. My gut feeling is that even though the Mac Pro and FCP are niche markets, they're still important to the company and they'll hang on to them, if only for prestige.

 

[Steve Sherrick]

If I were Steve Jobs, I'd want to rule the world. I'd want the content that is being streamed to iphones and ipads to have been created by talented people using Apple computers and Apple software. Even if it's a break-even proposition at best, the full circle appeal and knowing that Apple is involved throughout the entire process from creation to distribution would be enticing. By no means would I ignore the business side. If it meant losing a lot of money because of it, then sure, stick with iphones and let others handle the creating.

Personally, I'd like to see Apple actually figure out a way to leap frog others in the computing industry and present something that is years ahead of its time and appeals to professionals and consumers alike. They can tie in all their consumer products to it but let it be something that just blows people away, something no one sees coming. Push the computing industry years ahead of schedule. Basically, do what Jim did with RED, but go even further.

 

[Hrvoje Simic]

If I were Steve Jobs, I'd want to rule the world. I'd want the content that is being streamed to iphones and ipads to have been created by talented people using Apple computers and Apple software. Even if it's a break-even proposition at best, the full circle appeal and knowing that Apple is involved throughout the entire process from creation to distribution would be enticing. By no means would I ignore the business side. If it meant losing a lot of money because of it, then sure, stick with iphones and let others handle the creating.

Personally, I'd like to see Apple actually figure out a way to leap frog others in the computing industry and present something that is years ahead of its time and appeals to professionals and consumers alike. They can tie in all their consumer products to it but let it be something that just blows people away, something no one sees coming. Push the computing industry years ahead of schedule. Basically, do what Jim did with RED, but go even further.

And then there's the stock holder's meeting.

 

[Liam Hall]

I think it's time for someone to do to Apple what RED are currently doing to Sony, Arri et al.

Anyone know any visionary billionaires who want to shake up an industry that is flabby, arrogant and has lost it's way? It wouldn't take much; a decent communications strategy and respect for their customers would probably nail it...

 

[jake blackstone]

I don't know of any shop that's using the OS X software for full grading and/or finishing, it's primarily out there as an on-set tool and on occasion a learning tool (as well as being the basis for the Kompressor module, but that's another story). Do you? As far as I know, it doesn't support any hardware surfaces, which limits its usefulness as a production tool even if one did want to use it that way.

Also, I don't know why you're bringing up Film Master. The question I was replying to had to do specifically with Linux. My answer wasn't a review of grading systems, it was a comment on the significant penetration of Linux in our industry.

My post was to point out, that Linux is not the only OS, that is successfully used in post environment. That's why I had mentioned FilmMaster. As far as Baselight on OSX, I think my post itself explains why I had mentioned it. And, yes, OSX version supports JL Cooper, if need it, just like FLOS version.

 

[andrewhake]

Even the current form factor Mac Pro with some sort of direct and extremely high speed linking (be it Lightpeak or whatever) would solve most of this very niche groups problems. I think the biggest issue most people have with the current Mac Pro is the amount of PCI. Unless they redesign the internals completely (AGAIN), that is probably not going to happen, unless, the PSU gets dramatically smaller, the Optical Drives get completely removed (good riddance), OR. . .

You have an extremely simple connection option for clustering 2-4 Mac Pros together, all dynamically handling the available resources and managed by the OS. You do something like this and make open enough that other developers have access to it and you have a nice setup and all it took was some magical connector and a lot of complex software

Would also be quite effective as server replacements for people not needing an entire farm.

 

[andrewhake]

If I were Steve Jobs, I'd want to rule the world. I'd want the content that is being streamed to iphones and ipads to have been created by talented people using Apple computers and Apple software. Even if it's a break-even proposition at best, the full circle appeal and knowing that Apple is involved throughout the entire process from creation to distribution would be enticing. By no means would I ignore the business side. If it meant losing a lot of money because of it, then sure, stick with iphones and let others handle the creating.

Personally, I'd like to see Apple actually figure out a way to leap frog others in the computing industry and present something that is years ahead of its time and appeals to professionals and consumers alike. They can tie in all their consumer products to it but let it be something that just blows people away, something no one sees coming. Push the computing industry years ahead of schedule. Basically, do what Jim did with RED, but go even further.

Agreed. But it is hard to argue that they aren't already doing this, just not to the extent that you are discussing here. I agree with you I think they should take it further. Unfortunately as a publicly traded company you don't get to do whatever you want.

 

[Austin Glass]

I think it's time for someone to do to Apple what RED are currently doing to Sony, Arri et al.

Anyone know any visionary billionaires who want to shake up an industry that is flabby, arrogant and has lost it's way? It wouldn't take much; a decent communications strategy and respect for their customers would probably nail it...

I know of a guy who likes to do that sort of thing, but I think he's a little busy trying to ship cameras at the moment...