Some months ago, I reviewed Autodesk's render-on-the-cloud function, and found it less convenient and slower than rendering on my desktop. Some readers then complained that my model (which is one of the most complex ones included with AutoCAD samples files) was not big enough.
So I asked, how does one determine which models are sufficiently complex enough to pay extra to have rendered on the cloud, and which are small enough to render more quickly on the desktop?
Which leads to the next level of abstraction: if there are models too small to benefit from rendering on the cloud, what does this mean for those CAD firms wanting to move everything to the cloud?
If cloud proponents say that one of AutoCAD's largest, most complex 3D sample files (Oil Module.dwg at 2.3MB) should not be processed on the cloud, what does that say for the millions of other drawings that are smaller and less complex?
The big stumbling block of cloud proponents is the contradictions they expose by their arguments.
The cloud has its place for limited applications. This throwing an entire company's lot into the cloud is as silly as those who thought during the bubble of 1999-2000 that profits had no place in selling dog food over the Internet. Those who thought clearly back then knew that the Internet had its place for some things, but not everything.
As James Maeding wrote on blog.cadnauseam.com:
The people flocking to the cloud are consumers that need their phones and music backed up. The rest of us know the cloud as simply remote storage, and generally companies only use that if forced to by some client.
I don't regard file size as the criteria, it's time. If something takes an hour to render locally and it is quicker, or even the same but you can do something else while it happens, in the cloud that's the answer
Posted by: RobiNZ | Nov 22, 2012 at 11:15 PM
"... the cloud is simply a remote storage for phones and music ..."
A bold statement when published on a cloud blogging service :-)
Agree that makes no sense to perform cloud renderings of single small CAD models (at least until the whole CAD resides on the cloud).
Posted by: Vladimir Michl | Nov 22, 2012 at 11:47 PM
> A bold statement when published on a cloud blogging service :
Posted by: Ralph Grabowski | Nov 23, 2012 at 05:46 AM
I agree that time is the issue, but size is a rough proxy for time; what skews the time for rendering is the quality and materials. Nevertheless, cloud proponents are finally admitting that there are many cases for which all-cloud, all-the-time is less efficient than the desktop.
Posted by: Ralph Grabowski | Nov 23, 2012 at 05:48 AM
I think the flaw is applying desktop thinking to cloud technology. Many argue that large files are better suited to local storage but surely that comes down to how the application accesses data. I suspect my bank account is part of a vast database but I only get the tiny bit of data I need.
That said given today's tech I prefer the Skydrive/Dropbox model where the cloud is syncing data to local devices (and I can backup my own data locally) than pure cloud.
I wonder what proportion of those who don't "trust" cloud services use Internet Banking?
Posted by: RobiNZ | Nov 23, 2012 at 01:04 PM
The "but you trust online banking" analogy has been blown out of the water many times. Banks guarantee your data (a.k.a. cash) online, and so we trust them; the terms and conditions of cloud vendors guarantees no protection for your data online.
Posted by: Ralph Grabowski | Nov 23, 2012 at 01:57 PM
"The "but you trust online banking" analogy has been blown out of the water many times. Banks guarantee your data (a.k.a. cash) online, and so we trust them; the terms and conditions of cloud vendors guarantees no protection for your data online."
Ralph, I wonder how many time that is/needs to be said before people wake up just how true and important it is?
Posted by: R. Paul Waddington | Nov 23, 2012 at 03:16 PM
Dunno.
I expect to hear this discredited analogy next week at AU.
Posted by: Ralph Grabowski | Nov 23, 2012 at 03:25 PM
True so the issue is "terms and conditions" of use not cloud technology or application?
Posted by: RobiNZ | Nov 24, 2012 at 12:18 PM
RobiNZ; Yes, to a great degree terms & conditions of use is a major stumbling block to any thinking business person.
Posted by: R. Paul Waddington | Nov 24, 2012 at 04:05 PM
Robin, there are very significant unresolved issues with both aspects.
Posted by: Steve Johnson | Nov 25, 2012 at 07:44 PM
No Robin it is not just terms and conditions. It is what versions do we get to or have to use. It is security that can't be provided and that none of these cloud companies will indemnify us for when troubles arise. It is escalating costs of data caps over ISP's where none of these cloud companies control squat so they can't make valid proveable promises of cost cutting or overall efficiencies. It is how none of these cloud vendors, in particular I am speaking of cloud for cad here, will demonstrate in actual studies just how this will save money and time for customers over EXISTING infrastructure. It is about data ownership and things like the US government claiming that once it is in the cloud the creators do not own it http://yro.slashdot.org/story/12/11/02/1737219/us-government-you-dont-own-your-cloud-data-so-we-can-access-it-at-any-time
It is about the fact that the non disclosure agreements we have to sign with our customers these cloud people will not sign with us.
Please, do two things at once can only be done on the cloud? Really? Last time I checked I could assign one of my cores to my single core application and then use the other three to five cores for other things.
Another aspect is the trial balloon corporate speak and nonsense like referring to banking when we all know banking will stand behind it's product and the CAD world will not. It's in the fine print so go read it. These stories and conflicting comments of capabilities and promised products which change every week it seems lend no credibility either.
Here is a hint. I do not rent my house, I want to live in it and control it. I do not rent my cad because I want to control my expenses and my versions and my data. The very second I go to the cloud for any aspect of cad creation I cede control of so many things and really the primary goal of cad on the cloud companies is just this.
They want improved cash flow in ways that cant' be stopped. It has nothing to do with customer benefits and everything to do with pay to play. Prove me wrong here.
And finally it is this. With the combination of GPU CPU technology and mini super computers as cheap as under $10,000.00 why would anyone perceive a need to go to the cloud? While the GPU's are cranking on your renders your CPU's are doing cad at the same time behind your firewall and many times faster I dare say than any internet related thing.
So on the flip side of your comments can you give me concrete examples and reasons for just why I should want to go to the cloud for cad?
Posted by: Dave Ault | Nov 26, 2012 at 02:42 AM
Robin,
Read this fine article today at ZDNET. http://www.zdnet.com/who-should-own-the-keys-to-your-digital-assets-7000007881/
Why would any person want to do this to his company?
Posted by: Dave Ault | Nov 26, 2012 at 02:55 AM
" I do not rent my house, I want to live in it and control it. I do not rent my cad because I want to control my expenses and my versions and my data. The very second I go to the cloud for any aspect of cad creation I cede control of so many things and really the primary goal of cad on the cloud companies is just this."
Right I approve, this is the real problem (not just the security)......black clouds are forming, ready to pour their acid rain on the heads of the poor cad users....sigh!
Posted by: Andrè raison | Nov 28, 2012 at 02:16 AM