[Marcus Moore]

[Walter Soyka] “I’m not defining workstation based on expansion alone. The Tube will be a fantastic machine, but it will not be the best performer at anything. You have a CPU-bound application? A dual-processor workstation will be twice as fast. You have a GPU-bound application? A quad-GPU workstation will be twice as fast.”

I guess the question really is how many of the 850,000 workstations sold per quarter benefit/require the type of super-heavy CPU/GPU power you’re talking about. Can Apple soak up a majority of of the customers who would be well served by this MacPro, and leave the sliver of a sliver of the top-top tier of super-needy users to someone else. They already did this when they abandoned the xServe- realizing they just weren’t making any real dent in the server market.

[Marcus Moore]

I walked into my first full time job in 1996 as a Lightwave3D computer animator, and eventually migrated over to editorial as it more piqued my interest, and the garish and obvious use of 3D animation started to wain in corporate/commercial video.

So I understand the edict that there’s never enough horsepower for some applications.

But there’s definitely a schism that usually doesn’t get enough discussion. While editors at my level producing corporate/commercial/documentary are often called upon to be editor/graphics designer/animator. Ironically, the more “pro” an editor is, the more streamlined his function becomes. Someone cutting at the highest end of theatrical or broadcast television is likely never called upon to delve into anything beyond the most basic graphics or color correction- there are other people for that. Purely cutting HD or even 4K material at this point is not the computer intensive task. It’s the realm of FX and finishing that seems the most crucially impacted by computer performance- and will be for some time to come.

[Marcus Moore]

Via a friend, I did. It was very impressive.

It’s pretty amazing that the texturing can be done over animated sequences like that, though I would imagine that in actual production this would never happen, since animation would probably be tweaked well after character texturing had been completed. But it would definitely be helpful to texture a character in context, seeing where things break apart on the model.

And keep in mind that all happens without cloth IK, complex lighting or any of the other myriad of things that keep final render times on single frames of a movie like this up to about 10 hours.

If resolution was to level off, I think computing power could catch up to the needs of animators.

A lot of people seem to be skeptical about the benefits of 4K, but as someone who sits 1.5′ from a 27″ computer monitor in my office, and 6′ from a 92″ projection screen in my theatre, I see the limits of 1080p every day.

That said- we are definitely heading towards a law of diminishing returns. The number of people who will benefit from a 4K display is probably 10-20 % of the overall market. Going forward, the effective benefit of 8K is very likely to be exclusively contained to theatrical and other large screen venues.

This is all to say that I don’t think the upper threshold for Video editors won’t be a moving target forever. No one is producing in 4K now that doesn’t have to, and no one will be producing in 8K or 16K that don’t have to either. Human perception is the threshold, that that definitely has a definable edge, just like print does.

Advanced graphics will be like 3D animation, continuing to suck as much computer power as the complexity requires, but editing does have a horizon that I can envision- even if it’s 10-15 years away.

[Marcus Moore]

Exactly. I’m less interested in 4K for the work I do for it’s raw resolution (though yes, the ability for instant CU shots is wonderful) as i am for the latitude benefits RAW provides.

[Marcus Moore]

I agree with both you and Walter- however, it’s somewhat indisputable that over the last decade more and more professional users have seen their needs enveloped by “consumer class” machines.

I think it would take quite a bit of hubris on our part as Video Editors to think that eventual the wave of computing power won’t pass over us as well (especially what you consider what machines that do what we do cost 15 years ago).

While we like to think we sit atop the heap, above us is certainly 3D animation, which with ever more complex IK, ever higher textures, more computer intensive lighting and environmental simulations, and the same resolution and frame rate issues we deal with, will certainly be crying out for “specialized” machines long after we as video editors have seen our needs met.

[Marcus Moore]

The professional market has been ceding ground to “consumer” machines for some time. As things like iMacs have become more and more powerful over the last 10 year, they just keep scooping up more pros in different disciplines who’s need can now more than adequately be met by those devices.

I dare say; illustration, desktop publishing, programming, web design… are all being more than “adequately” served by laptops and desktops.

Even the lower to middle rungs of the A/V ladder can now be accomplished on these machines, and I think (on the video side, at least) the top tier keeps out of reach only because the top of the ladder keeping adding rungs, with higher resolutions and frame-rates coming into play.

But I think the top-top end is going to continue to get squeezed tighter and tighter. How powerful will and iMac be 10 years from now? CPU and GPU will be unquestionably faster, and if TB gets to it’s 100GB/s goal, I/O will cease to be an issue.

[Marcus Moore]

I’m wondering. What’s the use case for the HDMI 1.4 port on the MacPro?

With ACDs (Retina or not) being handled by the Thunderbolt2 ports, is the intention for the HDMI port to be used for Broadcast Monitoring? Rendering a separate Video I/O box unnecessary?

That is unless you’re using Resolve, in which case you’d still need a Blackmagic box, right?

[Marcus Moore]

Thanks for the explanation. I’ve read it thru twice now and believe i get the gist, as the amish say.

So it’s up to Apple to support external GPUs via TB. It will be interesting to see if they do, with big companies like Magma already building chassis they advertise can be switched to TB2 when it’s released.

With TB2 really just a rejiggered orientation of TB1’s data rate, I wonder what the next step for Thunderbolt is. Will it be a doubling- or is the next step to go right to the optical orientation with its 100GB/s data rate.

[Marcus Moore]

[Dan Stewart] “Now this one I’m prepared to bet on.. I think you’ll be more likely to upgrade your iphone.”

I dunno… the breakdown on Apple’s site makes it look very user accessible (at the very least for the RAM), and the GPU side there’s some VERY obvious screws in the design. The GPUs may not be a standard PCIe configuration, but as long as the MacPro design doesn’t change in a year, the next-gen AMD GPUs should fit- even if it’s something Apple recommends they do at a store.

I’m obviously not sure, but I don’t think the machine is as locked down as you do.

[Marcus Moore]

So everything that goes into a Cubix is sharing one 16x slot? That means you can really only add ONE heavy GPU worth of horsepower, whether it’s 1 beefy card, or several mid-range ones- right?

If that’s the case, then where is the power discrepancy between a current gen MacPro and the tower? Since the internal GPUs are presumably equivalent to 2 16x slots, and via TB2 you can use an external chassis for up to 8x cards.

I’m I nuts?