Thinking out (c)loud
I’ve got a couple of old friends who work for SolidWorks. Kishore Boyalakuntla was a Cosmos FEA application engineer for my neck of the woods when I was a reseller engineer in upstate NY. He has always been a great guy to deal with, and a great source of inspiration.
Mike Sabocheck was the SolidWorks territory technical mgr for my area during the same time. Mike also is a good friend, and has been a friend regardless of how I behave on my blog.
It just happened that I ran into these two guys in the Partner Pavillion today. We had a great discussion about the “cloud” stuff from the general session earlier. I hope that these guys associating with me in public does not cost them their careers.
In any discussion about SolidWorks, I usually function as the devil’s advocate, or skeptic or whatever you want to call me, and this conversation was no different. I just think any discussion needs a counterpoint to make it balaced.
Now this was just a little conversation between me and one guy who helps direct new products (Kishore) and one guy who helps sell it (Mike). We talked about the audience reaction to the cloud talk from the main stage today. No one was certain if the silence of the crowd was because they were underwhelmed, annoyed by more cloud talk, didn’t understand it, or simply asleep. I had to say that it was a little awkward to see Joe Dunne deliver the climactic punchline to the cloud initiative and hear only the echo of his voice in the big arena. It was that quiet.
Kishore seemed to have unfettered optimism for application of the cloud to engineering applications. He really offered a stirring rallying cry for reasons why it is so important and why customers want SolidWorks to be on the front of the technology curve. Everything he said, I had to agree with. There are a lot of great reasons to love the cloud.
As devil’s advocate, I had to point out that SolidWorks customers are not a monolithic single entity, but rather a wide range of users without a single voice who range from curmudgeon to techno optimist. Cloud application will appeal to a narrow section of these users.
Some of the advantages Kishore listed were that the IT overhead would be lower, the hardware requirements would be lower, the data management, tech support, and several other concerns are all minimized in this cloud scenario. Just so it’s clear here, we’re talking full blown SolidWorks on the cloud, not just some viewing application.
That’s great, the devil’s advocate replies, but what kind of business is going to put their data on a cloud server? From the stage, no one ever said where this server is going to be, or who is going to control it. In fact, Joe Dunne said something about having thousands of computers on this (hypothetical) server farm. Mike pointed out that immense server farms already exist, especially for companies like Amazon. Was the “cloud” they were working from actually just a computer behind the main stage? or were they live from Concord? Charles Bernard requested that people not use their iPhones to send messages while he demonstrated a 3dvia viewing app. So infrastructure is a concern. The cloud may be private, but the infrastructure is public, and for the foreseeable future is the weak link in this equation.
If you have a whole farm of professionally maintained servers behind you, you don’t have to worry about anything, including service packs, installation, or what kind of computer you are using.
Weren’t we all around during the dotcom bubble? Over hyped optimism in the internet left some useful remnants, but there was a lot of wasted effort for stuff that never came to be. Didn’t we all live through the mainframe stage of the computer age? everybody gets a dumb terminal? wasn’t the PC invented to give us independence rather than to be dependent on a centralized computer?
We debated it back and forth for a while. I think that we agreed that the long term result of this technology is unclear, but there is clearly some application for what SolidWorks is developing. For the short term, we already have some applications such as the stuff on SolidWorks Labs, and lately Project Butterfly.
(I’ll bet that SW showed this cloud stuff specifically to counter the recent Project Butterfly announcement from Autodesk. Jeff Ray did have at least one sarcastic jab at Autodesk during his talk. The difference between what SW showed and the Autodesk announcement is that Butterfly is on line NOW and works NOW, but the stuff we saw from the main stage is a semi-functional private prototype.)
Anyway, as three good natured debaters, we all gained something to think about, and stuff to speculate about. Speculation is fun, and it really isn’t all that dangerous as long as you don’t take it too seriously.
Ok, CSWP event tonight. Remote controlled race trucks. RRRRRrrrmmmmmMMM!!!!
Yes, I think it will be interesting to see where it goes. I know for certain things I like being able to depend on the cloud (email), but sometimes it’s nice to have the control. If they can address the security/reliability issues that many worry about, I think it could go far in helping people/businesses adopt the technology.
This sounds like it might be a great idea for my school. The issue of trying to run SW on old outdated computers is taken care of. Now the issue of cost to the school board, and if the cost includes giving my students access rights to work/practice from home have to be settled.
Software by the hour huh, what a deal for bean counters and no one else. I am with you Matt, this cloud hype is nothing but the latest group of fradulent promoters looking to relive the halcyon days of the dot com bubble where people would just mindlessly fork over the dough for nothing. Silence indeed as every rational individual was pondering if they would have to leave SWX if it becomes a choice of web and SWX or no web, no SWX
I definitely think it will be interesting to see where this all actually goes. For me there are a few things that I appreciate being in the cloud (email, calendar, tasks), that have made working and getting things done alot easier. However, they’ll have to address quite a few issues to get 3D CAD idea to stick. I’m all for CAD cloud stuff, but I am a bit skeptical as well.
Remember SolidWorks-on-a-stick back in ’08? Haven’t heard a lick about it since and it’s gone nowhere. Besides, it was constant crash-fest.
Matt,
I think for the case of FEA and rendering/animation the cloud currently makes perfect sense. If I could work locally and prep the model and then upload it for analysis or rendering/animation and have it send the results quickly back to me – I think I would actually be interested in paying for subscription. If I could gain access to computing resources that I don’t have or never would have to solve things that take a day or more traditionally – at a fraction of that time, I think that would be a great way for SolidWorks to start. If they could make that work and make it work well, then perhaps I would be interested in seeing what they could do well after that. I think they will need to take measured steps to convice the masses.
Reminds me of the old VAX terminal days of the late ’80s and early ’90s—central computing muscle with rather mindless displays/terminals for feedback/input. Actually a great idea for distributing highly expensive “intelligence” to many locations relatively inexpensively.
But the difference that bothers me is that the VAX systems typically ran under a single roof, or at least under a single entity (like an intranet). So all data and access were easily controlled by the creators of such data. In this proposed cloud solution, the data would necessarily need to reside (somewhere) in the cloud, right? What’s wrong with that? Additional failure points constructed between you and your data, plus the possibility of lost or stolen data (happens all the time with credit card numbers).
If, for whatever reason, your link to the cloud is broken, you’re out of business. Depending on who you are and where you work this may or may not be a big problem. I work from home in the mountains and have frequent interruptions of DSL service, requiring reconnecting to the Internet. If I’m working on the cloud, what happens to my model/assembly/drawing during disconnects? Am I able to keep unsaved changes and log in as soon as my Internet connection is restored?
The nice side of this is that if you’re traveling, you can use something light-weight to get work done—perhaps as light-weight as a little netbook! That would be great. No need to haul around a 12-pound laptop that costs 4X as much as the desktop equivalent anymore.
So I wonder—with good and bad aspects of cloud-based SolidWorks, could we have an option for either? Certainly some would value the good aspects of the cloud over the inherent risks involved. And if not fee-based, then everyone would need to pay for development and sustaining of the cloud infrastructure/support—as opposed to only those who make use of it. As for my particular needs, I’d like to opt out.
Computing in the cloud sucks. It depends on good reliable high speed connectivity. I sometimes work remotely. The best data rate that I can get is 4800 baud with a cost of $1.19 per minute. I could probably get one screen update every 20 minutes.
If solidworks would dehydrate the models to something more nearly the information content, these models would go from 8MB down to 100KB. I could then carry every version of every model I have ever seen on a single memory stick. That is light and small enough for me.
I want control over the software that I use. I would not risk a service pack change in the middle of a hot project. We have all seen models fail to rebuild with new versions.
The only good aspect of the cloud computing is that your model data is presumably stored in a highly redundant secure remote server that will be available everywhere all the time. I do not believe in Santa Claus.
Solidworks has a lot of bugs to fix before they expand into the cloud.
Matt,
I’m underwhelmed by the cloud initiative as well. While it would no doubt be useful if/when the vision is fully realized, I’d much rather SolidWorks put their engineering resources into edifying their current product and implementing key incremental feature requests rather than undertaking this major initiative. Of course, that’s not the type of direction that makes for a big marketing splash…
I agree with Pete above that some aspects of the software, such as analysis and even rendering, could be moved to the cloud in an advantageous way. Modeling, no way. We invest in fast workstations in order to make modeling as smooth as possible… it would take a lot more than cloud advantages to convince me to jeopardize that performance.
- rdo
Matt,
Don’t forget the other rhyming word “Crowd” as in “Crowd Sourcing”. I am sure it got lost in the heavy french accent that phrase was also dropped some where there. “Crowd Sourcing” CAD models is an interesting concept and more exciting that “Cloud Computing”. What is “Crowd Sourcing”…think Wikipedia.
Now how could I use “Crowd Sourcing”..Let’s say I am working a model and most of it is already done but I have complex surface for my model that I know you can do really quickly. I will put it out there asking some body to help to finish up a model and I will pay them a bag of peanuts.
CAD in the cloud is a bad idea. Even the most positive aspects can be inverted by small bugs or communications failures. Only people who do not use Solidworks would imagine that CAD in the cloud is a good idea. The cloud is only ok for my date book.
I do not want to pay for a server farm. I do not want to have my models at risk due to software updates. I do not trust Solidworks with my data. I do not want to have new bugs automatically added. I do not want to pay for ultra high bandwidth communication. I do not want to pay for the same old surfacing bugs year after year.
I want Solidworks to work. I think that they should plan on changing the name of SolidWorks to SolidFail.
@Joe Aggie
“I will pay them a bag of peanuts.” That much?
Hmmm, Alibre was born back in the Dot-bomb days as a “Cloud CAD” (OK, the terms were different back then, but the idea was that Alibre was going to use the Internet to make design collaboration much easier. IIRC, just the CAD data was stored remotely). Anyway, we’ve seen how that’s worked out.
I can see a local “cloud” working out (with high speed 100MB/1000MB networking), but I’m skeptical about remote (via cable/DSL/3G/whatever) clouds.