Creative Communities of the World Forums

The peer to peer support community for media production professionals.

Activity Forums Creative Community Conversations Nice thread on the future of the Mac Pro on the Videoguy’s forum

  • Walter Soyka

    January 2, 2013 at 3:55 pm

    [Jeremy Garchow] “I know you disagree with me, but you cannot go as fast on a Mac as you can on Windows.”

    Do we really disagree as much as this thread suggest?

    All I’m saying is that Apple sold some pretty sock-rocking systems, especially from 2006 to 2010ish. They were competitive performance systems in our market — the fastest CPUs, enough (but not the most) RAM, enough (but not the most PCIe), and the GPU didn’t matter all that much for most of our software yet.

    If there was a small performance difference on cross-platform software between a top-end Mac and identically-spec’ed PC, it was probably a worthwhile trade-off for most of us to have FCP on the same box.

    From 2010ish (with DaVinci Resolve on Mac being the first major application in our segment that could use multiple GPUs) through now (with no E5 Xeon), the Mac Pro has been outclassed.

    I hope they change that. FCPX is fast on a Sandy Bridge i7. It could be faster still on an E5 Xeon.

    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

  • Jeremy Garchow

    January 2, 2013 at 4:11 pm

    [Walter Soyka] “That’s the thing I disagree with you on. Just because Phi runs x86 doesn’t mean it runs your OS. It is a co-processor, not some kind of bolt-on processing power. OS system calls happen on the main system; these libraries are not available on the co-processor, which is essentially a separate computer within a computer.

    In other words, in some very important ways, A 16-core Xeon system plus a 50-core Phi is closer to a Xeon plus a GPU or to a Xeon networked to a little cluster than it is to a 66-core Xeon.”

    Yes, I understand, but the actual instructions, since they are the same as they are the Xeons on board, are very similar to write, unlike other instructions. Maybe I just can’t say it the way I mean it. I know that sticking a Phi into a Xeon isn’t a bolt on processor just like putting in an NVidia card won’t give you entire system CUDA availability.

    [Walter Soyka] “I think you could be on to something. I hope your prediction comes true.”

    I’m sure I’m wrong, but in my feeble brain, there has to be some reason to wait this long for a Xeon refresh from Apple. Phi could very well be a logical reason for it.

    On the other hand, maybe Apple is all in for i7/i5.

    I do see that a lot of people that have bought “the cheaper PC” buy core i7s with a decent GPU. It seems to be good enough for video work on the PC side, and perhaps it is on the Mac side as well.

    Jeremy

  • Jeremy Garchow

    January 2, 2013 at 4:18 pm

    [Walter Soyka] “Do we really disagree as much as this thread suggest?”

    Not at all.

  • Walter Soyka

    January 2, 2013 at 4:48 pm

    [Jeremy Garchow] “Yes, I understand, but the actual instructions, since they are the same as they are the Xeons on board, are very similar to write, unlike other instructions. Maybe I just can’t say it the way I mean it. I know that sticking a Phi into a Xeon isn’t a bolt on processor just like putting in an NVidia card won’t give you entire system CUDA availability.”

    Ok, I get it now. Sorry for my denseness.

    You are making a good point. With Phi, as a parallel programming developer, you don’t need to learn any of the special CUDA libraries or OpenCL APIs that you would have to learn to exploit other hardware.

    Have I gotten it now?

    This might be a two-edged sword. With CUDA, you have access to lots of libraries that simplify parallel development. For example, Ae’s 3D ray tracer runs on CUDA, not OpenCL, because NVIDIA has developed a CUDA-specific ray-tracing library called OptiX. Developers aren’t so different from us, and a library or API is a lot like an effects plugin: it may not offer anything that we couldn’t do from scratch if we had to — but why do it from scratch when someone has already built the core for you?

    [Jeremy Garchow] “there has to be some reason to wait this long for a Xeon refresh from Apple. Phi could very well be a logical reason for it.”

    Phi may be a logical reason for releasing a Mac Pro with Phi in 2013, but is it also a logical reason for not releasing a Mac Pro with E5s in 2012?

    Everyone says the worst part is not knowing. But you’ve pointed out time and again that we don’t really know with the other vendors, either.

    I think the worst part is not trusting. “I need a new workstation soon, but Apple sat the last round out. Will the 2013 Mac Pro have what I’m looking for, or should I think about an iMac or a PC?”

    [Jeremy Garchow] “I do see that a lot of people that have bought “the cheaper PC” buy core i7s with a decent GPU. It seems to be good enough for video work on the PC side, and perhaps it is on the Mac side as well. “

    I agree. Core i7/decent GPU is good enough for an awful lot of work. The iMac is a really appealing machine. The days of needing a tower/workstation/whatever for any kind of video work at all are behind us.

    Again, sorry for my denseness above, and thanks for taking the time to continue the conversation. Happy new year!

    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

  • Jeremy Garchow

    January 2, 2013 at 5:03 pm

    [Walter Soyka] “Have I gotten it now?”

    Correct. If you are already writing for parallel processing, and perhaps even offloading to Grand Central Dispatch, in theory, you won’t have to relearn anything. It’s fun to read article comments about Phi when it’s programer to programmer. There’s a lot of insight there.

    [Walter Soyka] “This might be a two-edged sword. With CUDA, you have access to lots of libraries that simplify parallel development. For example, Ae’s 3D ray tracer runs on CUDA, not OpenCL, because NVIDIA has developed a CUDA-specific ray-tracing library called OptiX. Developers aren’t so different from us, and a library or API is a lot like an effects plugin: it may not offer anything that we couldn’t do from scratch if we had to — but why do it from scratch when someone has already built the core for you?”

    Yes, as with most technology, Phi isn’t perfect and it won’t solve world hunger.

    [Walter Soyka] “Phi may be a logical reason for releasing a Mac Pro with Phi in 2013, but is it also a logical reason for not releasing a Mac Pro with E5s in 2012?”

    Not a good logical reason, unless I do more speculating and say that Apple was comfortable enough to take the risk and sit this round out. I could point to the supply chain, some sort of Thunderbolt ecosystem (assuming the next “MacPro” has Thunderbolt), or the ever present, the MacPro and anything like it, is truly dead. Long live the MacPro.

    [Walter Soyka] “I think the worst part is not trusting. “I need a new workstation soon, but Apple sat the last round out. Will the 2013 Mac Pro have what I’m looking for, or should I think about an iMac or a PC?””

    Yep. As I mentioned earlier on in this thread, all of this really boils down to the extinction of FCP7 and the Legend of FCP, which you have summed up in one word, “trust”. If FCP8 was here and had some 64 bit yadda-yaddas and Color v2.0 now with 5k, the hardware speculation would be less severe.

    [Walter Soyka] “Again, sorry for my denseness above, and thanks for taking the time to continue the conversation. Happy new year!”

    No no, thank you! If anything is dense, it would be the blatant optimism that I continue to espouse.

    Happy New Year!

  • Walter Soyka

    January 2, 2013 at 5:14 pm

    [Jeremy Garchow] “Yep. As I mentioned earlier on in this thread, all of this really boils down to the extinction of FCP7 and the Legend of FCP, which you have summed up in one word, “trust”. If FCP8 was here and had some 64 bit yadda-yaddas and Color v2.0 now with 5k, the hardware speculation would be less severe.”

    Conversely — do you think FCPX be less scary if 2011 and 2012 had also seen sock-rocking Mac Pro releases?

    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

  • Jeremy Garchow

    January 2, 2013 at 5:23 pm

    Q[Walter Soyka] “Conversely — do you think FCPX be less scary if 2011 and 2012 had also seen sock-rocking Mac Pro releases?”

    No. FCPX will still scare the people who don’t like it no matter how fast the machine.

    Upgrading to faster hardware is one thing (or dealing with slower hardware), but workflow is king.

  • Walter Soyka

    January 2, 2013 at 5:31 pm

    [Jeremy Garchow] “No. FCPX will still scare the people who don’t like it no matter how fast the machine. Upgrading to faster hardware is one thing (or dealing with slower hardware), but workflow is king.”

    I meant that more from a trust perspective than a capabilities perspective. I see editors both here and in real life rejecting FCPX, even where it fits (or wins) the workflow, because of the point of view (which I have certainly espoused) that Apple is not making decisions based on “our” needs. I wonder if that feeling could have been reduced by continuing to offer new hardware during this transition.

    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

  • Jeremy Garchow

    January 2, 2013 at 6:06 pm

    [Walter Soyka] “I wonder if that feeling could have been reduced by continuing to offer new hardware during this transition.”

    I’m sure it certainly could have reduced a small fraction of the stress that has been caused due the EOL of most of FCS3.

    People would still be extremely grumpy about it.

    While it might not look like exactly what you thought it might, you could replace today’s MacPro with an iMac and Thunderbolt peripherals and live to tell the tale.

    What you can’t replace as easily is FCS3.

    Replacing a MacPro and moving to an iMac with TBolt all round costs money and of course some time to get it installed, but everything basically, remains the same in regards to workflow. There’s no retraining. You can keep going on FCS3 while trying new software if you want.

    Replacing FCS3 with whatever, costs a lot more time.

    I’m not sure if a sizzle core would ease that pain.

    It would certainly help those users who didn’t use FCS3 in the first place.

    Jeremy

  • Steve Mcgarrigle

    March 25, 2013 at 4:03 pm

    “When was the last time anyone heard the words share and Apple used in the same sentence unless they were talking about their stock price?”

    Apple loves the word Share, they used it to replace all instances of the word ‘Export” in FCPX, Motion and Compressor. 🙂

Page 9 of 9

We use anonymous cookies to give you the best experience we can.
Our Privacy policy | GDPR Policy