-
My take on FCPX or Not
-
[Walter Soyka] “I don’t really see how a scalable desktop solves the PC industry’s woes.”
[Walter Soyka] “Video used to require a proper workstation. Consumer desktops lacked the throughput and processing power to push all those pixels around. That’s not true anymore.”
This is EXACTLY my point. This new computer can be an i7 multicore computer with the option for a very powerful GPU. The workstation features are added through Thunderbolt.
There’s not much of a market for workstations with lots of PCIe slots. Certainly a few companies like HP had the Z series. Apple is not likely to head in that direction.
Also facilities themselves are dealing with lower budgets in the economy. Workstations make less and less sense. Apple may solve this problem making what might be termed a “MacMiniPro” or a “headless iMac” The key difference is more CPU and GPU power. Expandability will be entirely through Thunderbolt.
A facility can be designed with high powered CPU/GPU locally (the workstation) and i/o and storage handled centrally through Thunderbolt.
The one man band who doesn’t want the limitation of the iMac monitor can buy one of these new boxes in the same price range as the iMac. They also have the option to get a more powerful box (cpu/gpu).
One box covering a broader market sector than the MacPro can lower the cost of production and, therefore, price. Yes commodification. Less differentiation drives down price of the base unit.
-
[Walter Soyka] “FCP was the linchpin holding Macs in professional post. With so many important apps being cross-platform, and with many also running on Linux, Apple needs to give the industry a reason to continue buying Mac computers.”
Yup. I’ve been looking at HPs and Dells this week. First time I’ve done that in about eight years.
-
[Walter Soyka] “My question for the forum is this: what do you think we can look for from Apple over the next 6 to 12 months to indicate if they are truly dedicated to the professional (or “complex workflow”) market or not?”
Thanks for starting a really interesting sub-thread, Walter. These hardware scenarios all sound pretty cool even if it’s all speculation right now.
I also think we need to think about software, since tight integration between hardware and software is one of Apple’s greatest strengths.
[Craig Seeman] “Apple’s motive though is to use FCPX to boost the sales of Macs so their potential ROI on the R&D is much greater even if it doesn’t come close to matching iOS devices.”
Agreed, but given the huge dominance of iOS hardware and software on Apple’s bottom line, how do you think iOS will impact Apple direction with the Mac OS platform moving forward?
I’m thinking specifically of Mac OS Lion as a first step in Apple merging the Mac OS and iOS platforms. Professionals and consumers have very different needs. Many UI decisions in Lion that seemed aimed at convergence and simplicity for consumers would not be tolerated in a professional setting. Some, like autosave seem downright dangerous.
What do you think Apple’s software direction says about their interest in professional workflow?
_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl -
[Craig Seeman] “There’s not much of a market for workstations with lots of PCIe slots. Certainly a few companies like HP had the Z series.”
Sure there is! That’s exactly why the Z series has been so successful. It’s a niche market, but those customers are willing to pay a healthy premium for the power they want in a single box.
My point is that very few video editors need a workstation-class machine anymore. A couple ThunderBolt ports on a high-end consumer-class machine will give them all the power and expansion they need.
For those that do actually need modern workstations, though, why limit yourself to 4x PCIe octopus expansion over ThunderBolt when you can have real 16x PCIe right in the box?
[Craig Seeman] “Apple is not likely to head in that direction.”
That’s my fear.
[Craig Seeman] “Also facilities themselves are dealing with lower budgets in the economy. Workstations make less and less sense.”
Workstations make less and less sense for editorial. You can reasonably edit on an i7 in a very small case.
It’s less reasonable to composite or work in 3D graphics without dual Xeons and lots of RAM. Once you’ve got those, plus a nice graphics card, plus a power supply and cooling to run it all, you end up with a big case again, with big, expensive components. Are you really going to reduce the size or cost of production drastically by leaving out a couple extra PCIe slots?
[Craig Seeman] “One box covering a broader market sector than the MacPro can lower the cost of production and, therefore, price. Yes commodification. Less differentiation drives down price of the base unit.”
The workstation I want — and am willing to pay for — is unlikely to appeal to the broader market sector. You won’t make my workstation small or cheap by removing 2 PCIe slots and replacing them with ThunderBolt ports. TB is not fast enough to expand to processors or GPGPUs, so I still need multiple big, hot, server-grade processors with multiple hot server-grade RAM slots, a couple of big, hot GPUs, big fans, and a big PSU.
That’s the whole point behind my concern over the future of the Mac Pro, and it’s definitely where I veer off from this conversation in terms of FCPX. Editorial used to need workstations, and the Mac Pro used to be a workstation. Now that editorial no longer requires a workstation, and now that Apple doesn’t sell any products that do, will the Mac Pro continue to be a workstation?
Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog – What I’m thinking when my workstation’s thinking
Creative Cow Forum Host: Live & Stage Events -
[David Lawrence] “Thanks for starting a really interesting sub-thread, Walter.”
Interesting? Maybe… Incendiary? Quite a bit more so than I intended!
[David Lawrence] “I also think we need to think about software, since tight integration between hardware and software is one of Apple’s greatest strengths… What do you think Apple’s software direction says about their interest in professional workflow?”
Does anyone see any evidence at this point that Apple thinks that there is even a difference between consumer workflow and professional workflow?
Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog – What I’m thinking when my workstation’s thinking
Creative Cow Forum Host: Live & Stage Events -
[Walter Soyka] “Does anyone see any evidence at this point that Apple thinks that there is even a difference between consumer workflow and professional workflow?”
I think John Gruber nailed it with this post quoting Ken Segall:
https://daringfireball.net/linked/2011/07/01/segall-pro
_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl -
[Walter Soyka] “Sure there is! That’s exactly why the Z series has been so successful. It’s a niche market, but those customers are willing to pay a healthy premium for the power they want in a single box.”
Niche market is the problem. Apple is not going to target the niche specifically. I do think they will have a comparably powerful core box in which through third parties you’ll add what you need. Apple’s design philosophy is single design to meet many needs. Apple’s core box would easily match CPU/GPU capabilities (as I would envision it). They just won’t design a box with lots of PCIe slots. People pay premium for the PCIe slots because there had been no viable alternative.
You need to rethink what makes a computer a “workstation.” CPU/GPU they can stick in anything with the power to handle that. Internal storage moves to external but just as fast or faster. Video I/O as external boxes is also very common these days.
If Apple designs a box which can range from i7 4 core to i7 12 core and a couple of GPU options with SSD for near instant on, that cover a much broader range of the market from upper end iMac to HPZ800.
The fact that HP desktops aren’t profitable and HPZ is says a lot about the computer industry. Apple will approach this differently. That’s in their DNA. iMac’s top out at i7 4 core and limited GPU potential and that’s an issue Apple will solve with this box.
[Walter Soyka] “For those that do actually need modern workstations, though, why limit yourself to 4x PCIe octopus expansion over ThunderBolt when you can have real 16x PCIe right in the box?”
Because if Apple makes a “commodity” box with the same power it will be less expensive and that’s going to lead purchases.
FCP sold not because it was “better” than Avid but was a lot less to buy a PowerMac and FCP than an Avid turnkey Media Composer system. Of course people still buy those systems and Avid nearly went under with the niche market dominance. Apple will do this with hardware. If they’ve learned anything from iOS devices it’s that mass production leads. 3rd party developers will sell the product just as the apps go a long way to sell iOS devices.
-
FCP X is like Charlie Brown at Halloween. “I got a rock.”
-
[Craig Seeman] “Understand that increasing sales doesn’t necessarily mean targeting the “high end” as a primary target but FCP1 was originally the “DV” editor. It’s popularity fostered 3rd party support and those 3rd parties help lift it into the higher markets. It’s those 3rd parties that will help determine whether FCPX sales more Macs. Apple has to supply them the APIs and hooks to do that.
“It was a DV editor when being a DV editor meant that it was a lossless editor for most. We were all moving our old betacam cameras to DVCam. Media100 and Avid had to ingest those via component and recompress them to 2:1 or uncompressed if you had a nice Avid. But in either case it was going to take a minor quality hit. And if you went back out to DVCam master, well, you’d take another compression hit because there was no direct DV connection. Media100 came out with a transcoding method and Avid trickled down a DV connection over the years. But in any case, my little G4 could with FCP (and EditDV for that matter) could actually produce higher quality content with a lossless workflow. Not to mention I could ingest Betacam via svideo and a camcorder, or via component with a targa card with v1.x of FCP. WITH RS-422 support and batch capture, etc. You could even get scopes if you knew how to do it. Yeah, no audio meters like some mention, but we were all importing from tape. So pretty simple to have my DVCam deck seeing the firewire connection for a perfect audio meter at all times. FCP 1 supported frame sizes larger than current 1080p. It had composite modes. It had precomps(nesting). FCP 1 was a VERY professional editor. ESPECIALLY compared to Media Composer. The only thing amateurish about FCP 1 was the price.
-
[Craig Seeman] “Niche market is the problem. Apple is not going to target the niche specifically. I do think they will have a comparably powerful core box in which through third parties you’ll add what you need. Apple’s design philosophy is single design to meet many needs.”
Isn’t that the main discussion here, though? We are talking about a small market of media professionals. This is niche that Apple used to cater to — a niche that is not satisfied with the single design that meets many needs — and a lot of the discussion in this forum has been dedicated to the question of whether they will continue to cater to that niche.
[Craig Seeman] “You need to rethink what makes a computer a “workstation.” CPU/GPU they can stick in anything with the power to handle that. Internal storage moves to external but just as fast or faster. Video I/O as external boxes is also very common these days. If Apple designs a box which can range from i7 4 core to i7 12 core and a couple of GPU options with SSD for near instant on, that cover a much broader range of the market from upper end iMac to HPZ800.”
I agree that storage and video I/O can reasonably be external, and do not a workstation make — that’s why I didn’t mention them.
You’re asking me to rethink what make a computer a workstation. I’ll ask you to rethink what kind of work you absolutely need to have a powerful, modern workstation for. Video editorial simply does not require a workstation anymore.
I think a modern workstation is defined by processors, memory, and GPU, as well as excellent integration and engineering. I don’t think you can just “stick [a CPU/GPU] in anything with the power to handle” it. There’s a difference between consumer-grade and professional-grade hardware, and I just don’t see how you can make it truly modular over a big market range with 4x PCIe over ThunderBolt holding it all together.
Core i7 won’t cut it — it doesn’t work in multi-socket configurations. You need to step up to Xeon, which adds cost. You need ECC RAM, because while a RAM error may be tolerable in a consumer machine, it’s a showstopper on a workstation. This adds cost. As we see more GPU co-processing with technologies like CUDA and OpenCL, those 16x PCIe slots will become more valuable and ThunderBolt 1.0 will not look like such a fat pipe.
All these expensive components draw serious power and require serious cooling, which requires space. None of these components are strictly necessary for straightforward video editorial anymore.
I agree with you that small, cheap, extensible computers are more powerful than ever before — but I disagree that you can make the core of a modern workstation small, cheap, and extensible.
We need a much, much faster interconnect than ThunderBolt to offer the kind of expansion I think you’re suggesting.
Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog – What I’m thinking when my workstation’s thinking
Creative Cow Forum Host: Live & Stage Events
Reply to this Discussion! Login or Sign Up