Quickie: Update on Autodesk and NEi

My goodness — the piece I posted yesterday about Autodesk acquiring certain asset of NEi Software really poked the bear! You’ve been commenting and emailing like crazy, which leads to this quick update:

  • No, I don’t know more that what I originally posted. There are still significant questions to be answered as to which products Autodesk plans to continue, how partner relationships will be handled, and which of your favorite NEi employees are transitioning to Autodesk.
  • All these years, I thought NEi Nastran was a flavor of Nastran. It’s not. Depending on who you are, that’s a good thing or a bad thing. Wrote one correspondent:

FWIW, I would argue that your post is somewhat misleading in that NEi’s product is not actually Nastran.  While it does use the word (acronym?) Nastran, it does not have the NASA/Cosmic Nastran pedigree that MSC, Siemens (and others in the past such as UAI and CSA) did.

Another wrote:

[NEi's competitors] relied on *old* ways. Companies and engineers choose NEi Nastran for its innovation and support.

Keep the comments and emails coming and we’ll all learn more together. But, please, keep it civil and constructive.

Print Friendly


4 Responses to Quickie: Update on Autodesk and NEi

  1. Monica Schnitger says:

    Your emails over the last few days clearly point out that there are different definition for what it means to be “Nastran”. How far back do the Nastran roots have to go to be legit? Is it a good thing or a bad thing for a solver to have add-ons? Are there any that don’t? I don’t have answers but it seems to me that the important point is what the software does today, whether whatever architectural decisions that were made in the past limit extensibility and accurate the simulation results are. But keep the comments coming!

  2. Dennis Nagy says:

    Anyone reading this piece posted by Monica should also take a look a her previous piece yesterday and the comments there, including mine. Some clarification was in order and you will find it there.

  3. Dennis Nagy says:

    Hi Monica,

    I re-read your comment above and realized that you had also asked a question about “legitimacy” of NASTRAN. My comment (to your other Autodesk/NEI post) was not meant to make any statement on “legitimacy” but just to point out the software architecture/detail differences between various vendors’ offerings that have used the NASTRAN label.

    The internals of the original NASA/COSMIC NASTRAN (vintage 1960s-1970s)are the basis for both MSC.NASTRAN and (because of the FTC case settlement in 2002) NX NASTRAN. MSC, however, had already changed/rewrote enough of those internals –e.g., the entire NASTRAN Executive System in the 1980s — that both MSC’s and Siemens PLM (SPLM)’s versions are not that much like the original NASA/COSMIC NASTRAN (or in fact like each other, because both MSC and SPLM have developed theirs further from the identical source code that existed in 2002). And, as I mentioned in my other post, NEi/NASTRAN is plug-compatible but different on the “inside.” Which NASTRAN is “better” (as a functioning software system) is only in the eye of the user/beholder.

    The marketplace reality is that “NASTRAN” has enough of a brand name (and that vanilla brand name is, as far as I know, still owned by NASA) that many vendor companies (MSC, CSAR, UAI, Noran/NEi, SPLM) have used it as a basis for building a software solution business since the early 1970s. The success (or lack thereof) of those businesses has depended much more on the quality of customer support, documentation, related pre/postprocessing, training, pricing, and customer loyalty. So that means, for example, that parts of the inner workings of NEi NASTRAN could be better than either MSC’s or SPLM’s NASTRANs even though not based originally on the same inner workings, as Katarina has pointed out in her comment about the nonlinear simulation capabilities, for example.

    And here is a final anecdote just to illustrate how counter-intuitive customer judgement can be: in the early 1990s, MSC implemented a “more accurate” 8-node solid element (with the same element name as the previous one) and released it in the newest version of MSC/NASTRAN at that time. We (I was at MSC at that time)quickly got a complaint/request from a major Japanese customer that they had tuned so much of their simulation workflow (e.g., criteria for judging accuracy and downstream usability of FEA simulation results) to tho “old” element results that we had thus inadvertently screwed up their major internal engineering simulation processes! They politely but firmly asked us to put in a “secret” switch just for them that would allow them to continue using the old, less-accurate version of the element.

    So go figure.

  4. This is quite interesting. What is intriguing is that I could not find a word of it on Autodesk’s website, despite them being a publicly listed company and the investment is presumably “non-trivial”. The NEI website also says nothing. If it weren’t for Katarina’s comment which makes no denial (and thus suggests tacit confirmation), plus a mysterious sales enquiry from a local Femap / NEI user who is now moving to Autodesk, I would have found the news difficult to believe. But it confirms what we have discovered by other means… the Algor solver was always more promise than delivery.
    And to save the moderator some time… I am a director/employee of a Siemens PLM reseller in Australia.