The company I work for use Catia extensively. But since Catia use cuda and since Dassault recommends Quadro, that is one piece of software that won't make it to the nMP any time soon. Here it is run on HP Zxxx serie workstation with Quadro M5000.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Is the new Mac Pro a failure?
- Thread starter ZombiePhysicist
- Start date
- Sort by reaction score
- Status
The company I work for use Catia extensively. But since Catia use cuda and since Dassault recommends Quadro, that is one piece of software that won't make it to the nMP any time soon. Here it is run on HP Zxxx serie workstation with Quadro M5000.
dassualt has openly said they aren't going to be supporting mac with their current software.
(ie- that is one (or two really) pieces of software that won't make it to the nMP.. ever)
If Rodney King had nMP and told them he was creative to the cops...what would they do?"can't we all just get along"? Pro? Semipros? Powerusers? Semipowerusers? Half-pro-power-semi-user?
Come on, it's christmas guys, and the new Star Wars is just around the corner
oh.. hey. it's lower case..
flat five ≠ Apt. #5
it's a flatted fifth
the blue note
the tone in between fa & sol if singing do,re,mi..
So. a week and a half or so into my goofy Windows system, and things are starting to click. I still need to sort out some of the ecosystem stuff and migrate over a bunch of VMs (so looking forward to consolidating those little boxes), but it's starting to feel like this will work for me. Still, it's somewhat awkward. I'm pretty sure that after a couple more weeks, however, it'll start to feel like home.
I got active on this thread and the MP group in particular, as I had some decisions to make on a refresh. Looks like they've now been made. But I believe that much of the info on this thread has helped me make a better-informed decision. And I still reserve the right to get back in this area, should Apple come out with a really nice upgrade that makes me want to sell my PC gear and get back on an MP. But for now, looks like I'll be back to occasional lurking status on the MP side.
But hey, I still love my little rMB!
I got active on this thread and the MP group in particular, as I had some decisions to make on a refresh. Looks like they've now been made. But I believe that much of the info on this thread has helped me make a better-informed decision. And I still reserve the right to get back in this area, should Apple come out with a really nice upgrade that makes me want to sell my PC gear and get back on an MP. But for now, looks like I'll be back to occasional lurking status on the MP side.
But hey, I still love my little rMB!
Since this appears to be for Beats Music, It could be for a series of music videos produced outside of what Apple normally does. Its most likely being produced using different software by a third party company and instead of converting it all to Apple products its using the native apps it was originally produced on. Might be part of a contractual requirement for all we know. Such as the upcoming Taylor Swift: The 1989 World Tour going to be shown by Apple.
Beats by Dr. Dre most likely have their own creative team in place using their own software & workflow and may have not made any transition yet, if any. Since they were only recently acquired by Apple not long ago.
Last edited:
I don't think any renderers in any packages I know of are written in .NET. Plugin support for all creative applications I know of isn't in .Net either.
Almost everything, even new code being written, has been written in Win32. There just hasn't been a reason to write creative software in .NET. .NET doesn't have a clear advantage for creative apps. Generally if it started in Win32, they're going to continue writing new code in Win32. .NET and Win32 don't necessarily mix easily.
Rhino's plugin support seems to be an exception, but they're the only plugin architecture I know of that's directly written on .NET.
I think you might be getting a little caught up in a point that really wasn't the point of what F5 was getting at, which is that software is migrating to more cross-platform-oriented development frameworks, largely spurred by mobile. .NET has components for developing those type of cross-platform applications.
ok. i see.. thanks
the software i'm most familiar with using uses .NET components in its bells&whistles features.. these features are now on mac, or coming soon, using the now open source .NET libraries.. i assumed (incorrectly) other developers worked in a similar manner.
It's true, while Dassault have been long pressed by customers to deliver an alternative for windows as OS platform, they have some project to Port Catia to Linux, both by user request and the EU request, no word on advancement but at some time they will release an Catia for linux, the issue es to translate literally Millios of lines of source code to another language friendly on Linux, most Catia code is on C# (not that difficult to port since Mono) but some code is MFC/VC++ which departs away GCC and is not easy to recode, which implies a long path just to have an 2-3 year old version of Catia fully running on Linux (assuming they get enough coders for that).
Actually any app could be migrated to any platform ( I hope Swift on Linux/ Windows as on OSX will help), the big barrier is MONEY, migrate catia, solidwoks its possible as long there is MARKET to may the migration cost, for CATIA, SOLIDWORKS the answer is NO, not enough market on OS/X or Linux users to justify the high expense migrating those large applications.
But What about that old Pig is AutoCAD, already we have Autocad 360 on OS/X and Linux, how? Autodesk is betting high on 3-D Printing revolution, also you can download a very capable Autodesk 123D for free as long do you do not use it for commercial purposes.
but Catia and SolidWORKS market is too vertical, I thing soon or later both will be platform agnostic as AutoCAD due the same reason, the popularization of 3-d prints and "in-house manufacturing".
At mid term CUDA is Doomed to disappear, support for OpenCL among the coders/IT community is strong , you'll see virtually no new project on CUDA or at least CUDA exclusive, software industry moved too quickly to adopt OpenCL as std de facto while other still use CUDA is due the costs implies to migrate to OpenCL, they prefer you to pay for nVidia than they to pay for OpenCL.
Yes/NO, they do not support Mac's as long are running OS/X but Mac on windows are supported, thats another history, as stated above with the Time both Catia an SolidWorks will adopt some platform agnostic offering as Autodesk is doing, and due the same reasons to take over the 3-D print "In House Fab" market.
When? I bet will take less than 3 Years from now to see at least the first 3-D Cad Tools from Dassault available on Win/Linux/OSX simultaneously.
That's not accurate.
.Net (C#, VB.net others) is like Java, its supposed you can write an app in .Net and run it efficiently everywhere (the purpose of .Net platform was to easy CPU migration, as these times when created MS and Intel beat on the Itanium as sucessor for the x86, and later MS repeated for Windows RT supposed to run on ARM as good it does on x86) the truth is that this have never been happen with perfect success or enough success to gain universal support from coders.
an windows app usually is coded on .Net platform since its large libraries easy coding in a significant manner, but here are few libraries due some specific requirement can't be coded on .Net, those are usualli the CODECs and CUDA/OpenCL extensions both requires either VC++/MFC or VC/WIN32(64) also Assembly is used, then those libreries are imported or linked to the .Net runtime to allow the performance or access impossible on .Net.
most codec not running on GPU on windows are coded on VC/Win32(64), for GPU ussually needs some Likner/Compiler specific for CUDA/OpenCL which implies to go awat VC/VC++ and of course by definition neither OpenCL/CUDA will ever support something like .Net Framework, same for OS/X while is easier to merge C and ObjectiveC code on xCode and CUDA or OpenCL could be coded right into the App, the same CUDA extensions and OpenCL are not running on the CPU only transfer the program control to the GPU, I'm not current on Swift but it seems more a lot like C#/.NetFramework cant judge it now.
Understood & agreed: they've not been part of Apple for long enough to have converted over...
...but by the same token, new job position announcements are also a forward-looking opportunity, so then just what does it mean to not even list FCPX as a required/desired skill?
In this context, the case is pretty reasonable that FCPX's absence implies that there are not even any future plans for them to contemplate FCPX..."ever".
-hh
Which is something you don't want in a performance sensitive application. That's why professional apps aren't written in .Net. You don't want the overhead of the runtime. Same reason you don't see professional applications written in Java.
You also just named a bunch of stuff that doesn't work in .Net, all of which are things professional applications use.
Swift is not at all like the .Net framework. It runs on the Obj-C runtime.
Not accurate, .Net is a framework, not a VM as java; you can write performance applications in .Net (ans ussualy are done in .Net), following an strategy of mixed code, this way you code the GUI and all the User-Iteration related on c#, and the routines that need high performance are coded on a separate .DLL library in VC++ usually in Win64 or MFC, then inside .Net you "invoke" these routines and you get your data processed at top speed with minimal GUI coding.
From what I've read Swift is more line a .Net app with he framework hot-linked as an dll library instead to use a sort of run-time compiler.
Swift is not like a .net app at all. Swift 2's foundation types bridge directly to Objective C Foundation types, this however is all changing with Swift 3-ish as the version currently released for linux doesn't bridge to anything.
My transition is now finally over. Brought my 6.1 to it's new owner and got the 5.1. GPU (GTX 980) arrived this morning , inserted couple of SSDs and put in 4 harddrives. Performance versus my 6.1 (6C,32GB,D700) is way higher than before. Best choice I've ever made so far.
Same thing on different words; for those familiar with MS programming, this is like to program on vc++/mfc but instead code in vc++ you code on Swift and the mfc is build on objective-c.
Ok, I'm planning to sell my nMP asap the 7.1 model is available to order... My total expense would be higher than yours but I'll get 8-10 cores E5v4 and Fiji based GPUs about 2x faster than GTX980Ti and still fit inside my burden.
Last edited:
You still don't have CUDA and internal Expansion - which could be perfectly fine depending on your usecase. However, the performance boost I gained from upgrading from the 6core to 2x6core in the cMP is a well fitting upgrade to my workflow, especially when looking at 4-6K footage.
I was thinking the same thing, though I may selfishly keep the 6,1 and use it as an OS X / Xcode server, it's power usage is crazy low at idle.
As he already has a 6,1 then isn't giving anything up here, if goes for a 7,1 which is expected will likely come with AMD GPU's again.
In my personal environment CUDA doesn't matter anymore, and our workkflow relies on external(thunderbolt)/Networked Storage, If you cant evade CUDA (or get OpenCL) in your workflow I understand, nobody cant stop to produce just to support apple's game, while I consider OpenCL will dominate and replace CUDA for good..
I'm really satisfied with my "new" 5.1. Since it's the complete maxed out version CPU wise and I can offload very intensive tasks to my Windows workstations, the 5.1 should be a pretty longlasting experince. It's incredible in my book that a 5 year old machine still holds (partly) up in today's world and through upgrades via PCIe it will be a long time before I'm finally going away from the cMP. Whatever comes after the 7.1 or even 8.1, I don't care right now.
- Status