Have you heard? Ansys to aquire FLUENT!!!

All you guys better start learning Fluent I guess, CFX will go the way of the dodo!!

Not likely.

Of course they will kill off CFX, why do the need two codes, they will strip CFX of the best bits then convert the users from one to the other (tempting them with free training or some such). It doesnt really matter to ansys, they bought CFX for so little money they wont have lost much, they will just have a massive combined market share.

Where is this news published? I didn't find any announcements on either fluent's or ansys's websites. Confirmation?

It is official, check out www.cfdreview.com

Yes, it's the begining of the end to CFX. If CFX is good enough, Ansys would not need to spend such a huge amount of money to acquire Fluent. $20 millions against $565 millions, which one would win?

But what about the user interface? And turbomachinery-specific capabilities? I hope there is some significant cross-pollination there.

They will probably rip the bits that CFX is good at (i.e. the turbomachinert stuff) out and integrate it within fluent, I imagine most of the bits that are taken will be more solver side and possibly meshing side. I doubt very much that they will retain the gui, they have one that does the job why have another?

http://today.reuters.com/stocks/Over...?symbol=ANSS.O

Check Ansys' stock price

You clearly have no understanding of software architecture. It is not practial to "rip useful bits out" and put them in Fluent. It simply does not work that way. The two solvers are fundamentally different as well, so it would be an impossible task anyways.

If any stripping is done it will be the other way around. Useful models from Fluent are reimplemented in the CFX solver, which is a better core solver. Maybe the whole thing is rebranded or something as well, who knows.

It's all speculation anyways.

Dan.

Is that you Williams? biatch. While the CFX technology might be a bit better (not quantum leaps ahead like you guys like to think) it'll cost way more to convert the Fluent user base than vice versa.

Sorry, but I hear there's some good MTB'ing in Vermont

Yep, you fingered me man.... I agree, not quantam leaps ahead.

The bottom line is that any migration of models/code from CFX to Fluent or Fluent to CFX will involve a rewrite. CFX is coupled element-vertex based, Fluent is segregated cell-centered. CFX is written in FORTRAN, Fluent is written in C or C++ or whatever. It's a very difficult software engineering problem.

I don't see that from a core solver standpoint that the CFX into Fluent route makes any sense, you loose the benefits of all the coupling we do. Many of the models in Fluent can easily be reimplemented, some of course may be harder though...

At any rate, this "merger" "aquisition" whatever you want to call it, will give us a chance to really get into the guts of things and see what can realistically be done.

Neale

Word. Totally agree. I love these "oh yeah, they'll just cut'n'paste some subroutines from one, a couple from the other and then badda bing badda boom we'll have a nice new software in 6mths". More likely to pick a platform and use the developers expertise to add to it. Still takes a lot of time though.

If history is anything to go by the end user won't see anything new for a couple of years. Heck the sale won't even be finalised until the summer.

BTW I moved to Montreal and went back to being a roadie homo. Riding a bunch of track too... even came down to London to ride the new velodrome there.

Yes, that's right. It will be status quo for a while yet. CFD customers tend to not like revolutionary change.

Things are still pretty good with me. Started with a decent girlfriend, then wife and house, now a kid, la la la life goes on. Get on your mountain bike and go over to Bromont sometime in the summer. You will be satisfied. I ride there now and then.

Bromont rocks. That's where the velodrome is for me, but I did ride the MTB there once. It was ace, but a bit limited because they were blasting some ski hills at the time.

Now you've procreated you're equalling Shannon on the Darwin scale according to Jerry's theory. Me too

Later

With our combined efforts maybe we can eventually beat down the effects of Shannon on the species.

I didn't get a bike pass at Bromont last summer because of the construction. I heard many of the decent DH runs were closed as a result. Still though, an excellent riding destination, much better than anything around here.

Hey Neale/Dan,

well if if solver architecture is an issue...parlez vouz fluent....

CFX-5, as you probably know struggled with sover archetecture infastructure for many releases and re-writes.....took years....lost all kinds of potential customers...hence the 23/~550k ratio on purchase price!!

For adding capabilities for multiphase problems for example, CFX did full re-writes.....fluent added an index to the domains structure??? Yep CFX all coupled up can do some problems better....but look how long it took to get it to market!!!

Core solver technology for constant or near constant propery flows...yes much better solver with the pressure based coupled solver techonlogy for steady problems....no argument...other stuff....think again...

just my oppinion!!

Regards......guess ;-)

Yes, the CFX-5 solver was written (not rewritten), once, from the ground up. So were the GUIs, i.e. written once from the ground up. This project mainly resulted from the required product transition of CFX-4/TASCflow to CFX-5 as well as needing to support unstructured grids. CFX-4 and TASCflow were also becoming unmaintainable so a new development platform was needed. What is there now in CFX-5 is a completely maintainable platform.

In the end though who cares how long the software took to come into it's own. The important thing is how it works today. Dwelling in the past gets you nowhere.

I don't understand what your comment about constant property flows or "other stuff" is supposed to mean. Pretty nebulous. These days I'd put CFX up against anybody, any day, anytime.

I definitely agree with you about the original price ratio. When CFX was purchased by ANSYS it was still not making money. AEA spent a lot of dollars funding the writing of CFX-5 and sold CFX to ANSYS at a bargain.

I'm sorry you simply jumped ship before things got really interesting.

I think Ansys will bild up a great puzzle with the useful features of Fluent an CFX. Ansys already did this with CFX and Workbench. Haven't you ever heard about "modules"? For example there will be a module with the CFX fetures for Turbomachinery and a Fluent module for combustion. There won't be the problem of the integration of the two solvers. The big match is to understand if the name "CFX" will persist or not. In my opinion the number of clients that Fluent has let immagine that CFX will disappear (externally!). Have a nice day!

Mica