Isn't asking Pixar to demo and give reactions to a machine akin to a father asking his favorite child what the child thinks of him?
I'd say what the computer was actually able to do during the demo spoke for itself.
Isn't asking Pixar to demo and give reactions to a machine akin to a father asking his favorite child what the child thinks of him?
5870s can run OpenCL. On Windows it was always CUDA based, or at least required an NVidia card.
What software are pixar using? Maya? Really excited to see the video.
I'd say what the computer was actually able to do during the demo spoke for itself.
I'd say what the computer was actually able to do during the demo spoke for itself.
This.
I don't think a majority of the people who were unimpressed realized what exactly they were looking at and how much information was being hauled around the screen.
On top of that, didn't they say something about the software being ported to OS X/OpenCL 8 days before the demo?
I think we can all agree the demo was impressive, but again, this was a software demo.
I think we can all agree the demo was impressive, but again, this was a software demo.
Until we see the Mac Pro put through some tests against other hardware then I wouldn't put too much stock in it.
I think this viewpoint is a shame. Moving over 20 GB of data around in a 3D shaded environment in real time in a software that was ported in 6 weeks and probably never optimized for the hardware is a shocking demonstration. To me, anyway. I'm not sure how anyone who understands what is being demonstrated can say that this is a software only demonstration. You can't have this kind of a demo without some serious firepower under the hood.
I'm not sure what people are expecting to accomplish by benching it against other machines. I'm guessing it'll run just like any other 12 core Sandy Bridge Xeon with dual FireGLs and a gigabit PCI Express SSD.
What surprise are you exactly expecting to see when benching it against other hardware?
I think this viewpoint is a shame. Moving over 20 GB of data around in a 3D shaded environment in real time in a software that was ported in 6 weeks and probably never optimized for the hardware is a shocking demonstration. To me, anyway. I'm not sure how anyone who understands what is being demonstrated can say that this is a software only demonstration. You can't have this kind of a demo without some serious firepower under the hood.
I'm not sure what people are expecting to accomplish by benching it against other machines. I'm guessing it'll run just like any other 12 core Sandy Bridge Xeon with dual FireGLs and a gigabit PCI Express SSD.
What surprise are you exactly expecting to see when benching it against other hardware?
On top of that, didn't they say something about the software being ported to OS X/OpenCL 8 days before the demo?
A shame? Really? How can you call it anything other than a software demo, especially since that's exactly what it was? Of course the hardware is powerful. We already know that based on the specs Apple has listed. But unless you see this against some other hardware you cannot possibly know what to contribute to the hardware or software.
Uh, how else are you going to value it? On its size and looks alone? You're saying there's nothing to gain from comparing to competing products?
Emotions have taken over and all sense of scientific comparison are gone.
If you were shooting a commercial for the 2014 Camaro and wanted to make sure everyone knew it was X% faster than 2013 Camaro would you:
A. Show a 2014 Camaro streaking around a racetrack by itself.
B. Show a 2014 Camaro showing it's tailights to a 2013 Camaro as it whips past in full throttle glory on the same racetrack?
I could shoot a 1985 Hyundai zipping around corners and it would look fast. It's only when you have something else to compare it to that you can judge actual speed.
If you genuinely want to impress people with how fast something is, you don't ONLY TEST 1 thing.
Imagine how pointless Barefeats.com's tests would be if he only gave you one item tested with no yardstick to measure against.
"The New Mac Pro can compute at 1.8 Gigawatts"
Huh?
This demo was for Mari users. Mari users (if you haven't been paying attention to the thread) are impressed because they know how fast Mari normally runs. They don't need to see a side by side, they're already seeing a side by side with their own machines.
You don't normally see car commercials with another car, because again, you already know how fast your car goes.
It would be like if Barefeats only posted the scores of the new Mac Pro, while you still had access to the scores of every single prior benchmark.
This requirement for a bakeoff is totally artificial. Strangely enough this artificial requirement is coming from someone who's side business might be broken by the new Mac Pro. Huh. Fancy that.
Oh no, what I meant was the a 5870 probably could not produce this sort of experience. So I'd still expect this machine to be way faster than the current.
I'm just skeptical when they don't provide any amount of repeatable measurement or test.
This demo was for Mari users.
I don't think so, the demo time was spent showing dumbed down 3D animation work which Mari users certainly know. Additionally I asked a friend who works in this industry and he says that Mari is for a small, specialized group of people (who charge a LOT). The bread and butter is elsewhere. In other words Apple isn't going to sell a lot of boxes for Mari.
The story is (as I read it) that even Tim was impressed with Mari on the Pro, so it seems reasonable that they just liked it because it was a good story about video work that needs horsepower, that would show well (because it's animation and so easy to grasp).
It was. The entire demo was billed as a story of a developer porting a specialty app to the Mac. The Mac Pro was a side bit. The Mac Pro thing was barely mentioned on the conference schedule.
This was a conference for developers, remember? None of them were interested in the app, only the porting process. I can tell you no one in the room was complaining that they didn't benchmark the Mac Pro because that's not what this event was expected to be.
The launch event will likely be full of benchmarks. But this wasn't a launch event. It was a developer aside focused on porting software, with a demo on a machine that won't even be released for months. This video isn't even supposed to be available to non developers.
I think that's exactly what we're trying to say. It was an impressive demo that happened to include the Mac Pro. It's the reaction of some people in these forums that seem to be using this as some sort of Mac Pro hardware demo instead. We're just saying, hold up, don't dwell on this too much.
This was a conference for developers, remember? None of them were interested in the app, only the porting process. I can tell you no one in the room was complaining that they didn't benchmark the Mac Pro because that's not what this event was expected to be.