This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

(Long term) Availability - TM4C series?!

Other Parts Discussed in Thread: TM4C1290NCPDT

We are about to re-design ~15 boards with LM3S Chips :-( Because of the compatibility of the Tivaware with the Stellarisware the TM4C is obviously (and regrettably) our best choice. Even tough TI was not able to create pin compatible replacements. (it's not just about PCB design => EMC tests, test equipment etc.!)

The expected overall cost for the redesign is more than 200 thousand USD. As you can imagine we are not keen on doing this again in the next years.

LM3S had been promoted as a long term available product to us. Now we hear the same phrases from our distributors and TI for the TM4C.

But to our surprise we can hardly find TM4C derivates on the market. For instance the choosen TM4C1290NCPDT is only available with long lead times. We take this as a sign of backing the wrong horse again...?!?

What are TI's plans with the TM4C series? What is the opinion of the community?

Thanks.

  • I'm curious about TI's answer too.
  • One of the first places I check when picking a micro is digikey (mouser would work as well). I look to see if they keep stock and am wary of parts that do not have stock since that indicates they are not widely used. Performing that check on TM4C yields 80 parts with stock, a few marked as non-stocked items that they do not regularly keep stock on.

    If an individual part has long lead times that usually indicates that it's new, there is little demand, there is huge demand (perhaps from one or two customers) that is consuming stock or it's on its way out. I view that latter two cases as problematic, the first I step carefully because product introductions can slip.

    If you really need assurances you should talk to your TI rep to see what real assurance they can give you. The longer period of time you need assurance for the harder it will be to get. Also find out what their end of life (EOL) policy is and how to be notified. Bear in mind that sometimes you will see chips change to include bug fixes or obsoleting in favour of a more capable sibling in the product family, you have the possibility of needing to recertify in that case as well depending on what you are certifying to. In the end you need to worry about this not only for the processor but for all parts on the board.

    As a cautionary tale let me remind you of the behaviour of a big chip supplier. They had been a supplier of embedded chips for a long time, having started with micros as one of their main products. Having drifted away from the market they were trying to get back in and made a big deal of promising (IIRC) a minimum of 5 years from the product introduction to end of life. A year or so later they made a big introduction of a processor for the embedded market, the actual production ended up slipping something like 4 years before any volume was available. Shortly after customers had designs made they announced they would stop making the processor in a year. When questions and complaints arrived they pointed out it was more than 5 years since they announced the product! Mind you that same customer kept selling an embedded processor for I estimate 15 years after they had stopped further maintenance, finally stopping production on all variants when they finally closed the fab.

    In the end you have to rely on judgement and reputation. There is one manufacturer I will only specify as a second source unless really pushed because of delivery issues.

    Robert
  • Like you - our firm (and our key clients) were nearly "killed" (financially, at least) by the LM3S "handling." (to be very generous)

    My take is bit different than that - thus far - presented:

    Some way/how find a device (employed in volume) by a "giant." In our case - we often consult or contract for such firms - parts "winning their sockets" (often in EAU > 250K) have best likelihood of surviving, "Culling of the herd!"   And also - there's a clear double benefit - we're paid to learn (and master) that chip - then may use it again (always non-competing) and we often can "nego" best price from the "giant."

    Another means - believe "Lloyds of London" (or similar) most always are willing to insure "uncertain outcomes." Now I know this is beyond many/most (i.e. all) here - but my firm does such when "hedging international currency fluctuations" - may merit your investigation for your circumstance, too. (Hollywood actresses insure (here I do not lie/exaggerate) special body parts!))

    Of course - if you are skilled (and lucky) in forecasting - you may buy & "stockpile"- sufficient inventory for some "reasonable" future...

  • cb1- said:
    Some way/how find a device (employed in volume" by a "giant."

    A company I've been with in the past has done that to their detriment, luckily it did not require a complete redesign.  Two things can happen

    1. The part can go on allocation, the giant gets preference
    2. The giant ceases to use the part and it disappears from the catalog

    If you can find a part used by many medium sized players both those side effects are reduced since the loss of a single buyer doesn't put the whole line in jeopardy or fixate the suppliers attention.  Hard to be sure you have figured that out though.

    cb1- said:
    Another means - believe "Lloyds of London" (or similar) most always are willing to insure "uncertain outcomes."

    A variation on disaster insurance, presumably applied to all/most of the BOM.  Something to buy you time to re-design. Hmmm....

    You can also, with great care, look at the grey market for manufacturers looking to get rid of overstock they bought on a last time buy. Doesn't prevent the need to redesign but it does buy time.

    Robert

  • I'd be shocked if o.p. "anticipated" depth & breadth of (non-vendor) response.    And really - what can (this/any) vendor possibly say - or add?

    While it may happen that the "giant" turns to another device - it is not "normal/customary" for such change to happen often - and never quickly!    And - most responsible "giants" buy forward - preventing the mangling of their production lines.   (thus they're committed - even more - to their chosen device)

    Repeatedly - our firm has gained superb price discounts - by nego'ing component buys directly from the "giant" - thus taking advantage of their significant buying power.   (always unavailable to we mortals)

    And - should the (unlikely) scenario of part, "going quiet" occur - as we always strive to remain on "best terms" w/clients (just as we do here) we'll likely be among, "First to know" - and thus enjoy the option of, "Stocking up" and/or "Switching to the "Next Choice" of yon giant"...

    No technique nor method is perfect - yet I'd bet heavily that such, "shadowing of a giant" - as described here - far exceeds alternatives!

    And - any "excuse" to meet & transact w/a "giant" - most always proves worthwhile - completely missed - by any/all "lesser methods!"    Q.E.D.

  • I am afraid that Texas give up of the Cortex-M platform. Two years without a new Tiva. In this time ST annoucned the STM32F429, STM32F334, STM32L4, STM32F7 with cortex-M7. Freescale announced the MK64F, MKV3x,MKV4x, MKV5x with cortex-M7. Atmel released the SAME7x, SAMV7x with  300MHZ cortex-M7 and with ethernet and packege option from  64 pins LQFP. Etc..

    I still have hope that Texas will soon announce new and competitive cortex-Mx for TIVA family of microcontrollers, because Texas still have by far  the best development support.

    Ari.

     

  • Ari Mendes dos Santos said:
    because Texas still have by far  the best development support.

    And - according to rapidly rising poster Luis - best LPad T-Shirts...    Although my small firm - gifted often - find 2 cycles thru the wash & "C'est fini..."

    Say Ari - its been months (or at least few weeks) since you've discovered & reported here (another) "Special/Custom" MCU - not fit for we miscreants...

  • I am afraid that Texas give up of the Cortex-M platform.

    Seems not so. A few days ago TI announced the "new" MSP432, which is incompatible with the 430, but has a Cortex M4 core ...

    While it looks like a commitment to the Cortex M platform, the TI's strategy is not clear to me. The microcontroller portfolio has a lot of gaps, inviting customers to "jump ship" and board at competitors...

  • it would be nice to know when TI would be releasing Cortex M7.

  • Amalinda said:
    ...nice to know when TI would be releasing Cortex M7.

    Might that occur just after "PF0/PD7" and (always helpful) 0Ω plague (resistors on LPads) are (properly) & finally resolved?

  • swtAT said:

    The expected overall cost for the redesign is more than 200 thousand USD. As you can imagine we are not keen on doing this again in the next years.

     It seems same for me.. It's time to wake up this dormant question, time over 3 mount has elapsed no answer at all, so what about competitor do for designer with assurance of product life?

    www.st.com/.../SC1169

    swtAT said:
    But to our surprise we can hardly find TM4C derivates on the market. For instance the choosen TM4C1290NCPDT is only available with long lead times. We take this as a sign of backing the wrong horse again...?!?

     IMHO I feel as I got again the wrong horse..

  • Roberto Romano said:
    time over 3 mount has elapsed no answer at all

    Roberto - while we "feel your pain" there has been an answer.    Search under TI's "David Maples" - he provided a very fair & open response - and I believe w/in the 90 day window you cite.

    Business experience & sense rises in value in this instance.    "Buying forward" - either in currency, commodity, or IC - is a well-known "hedge" against unwanted and/or unexpected market movements.   

    We've long done such - sometimes look like genius - other times - not so much - yet we (and our clients) are able to (somewhat) defend against shortfalls or other situational unpleasantness...   Planning - even for the less likely - trumps "hand-wringing."

  • Often the commitments are rather less than they appear, to quote

    In case of significant volume decrease, technology or manufacturing changes, a switch to a comparable product, another technology or a different manufacturing facility could be decided by  *supplier*

     

    And comparable leaves a lot of room for changes, not requiring any of pin compatible, same peripherals or even same power supply. Basically the commitment boils down to if anything changes we reserve the right to stop supplying the chip.

    I remember another commitment to maintain long term supply that ended up amounting to "here is the production version of chip and now is the time to place your last time buy". Thankfully, it's not one I was using.

     

    Robert

  • cb1- said:
    Business experience & sense rises in value in this instance.    "Buying forward" - either in currency, commodity, or IC - is a well-known "hedge" against unwanted and/or unexpected market movements.   

     Thats a good Idea, commercial goal sell now than do a production plan over time... This is good for vertical market, averaging for mixed market and WORST for horizontal long term market done by a small firm.

     So this question is use or not TI? Why too many are asking if I use TI component on design and offer to help rework all design? MMMhhhh is some conspiration, political reason or something we cannot know?

  • Robert Adsett said:

    I remember another commitment to maintain long term supply that ended up amounting to "here is the production version of chip and now is the time to place your last time buy". Thankfully, it's not one I was using.

     Hi Robert, thank a lot for all precisation about, I am happy for you but I feel like I got again on wrong side, it is not?

  • Customers always ask for certainty.

    I remember a potential customer asking for a 20 year guarantee of parts availability from the point that they ceased production of the device it was going into.

    It's like the perpetual requests for absolute code security.

    Robert
  • I don't think you have any more assurance from one supplier than another. You do have a colourful image and a promise nothing statement from one and the appraisal (that I think is honest) pointed out by cb1 from TI.

    You could try to negotiate security from the semiconductor manufacturer's but you would need really large volumes I expect to have any chance of having that kind of conversation. cb1's offered a practical alternative and it should be noted that precaution extends to you entire BOM, not just the micro's. Of course for some part's it may be that the needed precautions are simpler but the risks are broadly the same.

    Sometimes the most sensible risk mitigation plan is simply to price in a redesign.

    Robert
  • Robert Adsett said:
    cb1's offered a practical alternative and it should be noted that precaution extends to your entire BOM

    Robert - at some point my keyboard will "wear away" hitting the "Like" button for your many posts!

    May I note a general agreement for your quoted statement above - with the exclusion of, "jelly bean" parts/components - available (always) from multiple sources - and at "clustered" prices & lead-times.   Thus - "buying forward" of the mass BOM may not be (really) required.   (only the most critical ones usually receive such "protection.")

    To Robert's and Roberto's point - "Can a promise be made to, supply forever" - might the same extend to, "Marriage, or business organizations, or the (oft unspoken) contract between parent & children?"    (love what law-school does to you...)

    Renewable contracts - rather than ones, "Cast in stone" seem often highly advantaged.   People and organizations often do not, "Grow at the same rate and/or together!"   What then?   In contrast - time-limited contracts intend to place "both parties" on their "best behavior" - so that such contract has the best likelihood of "being renewed!"

  • cb1- said:
    with the exclusion of, "jelly bean" parts/components - available (always) from multiple sources - and at "clustered" prices & lead-times.

    In general I would agree with this, with two small notes

    1. If you are doing certain kinds of certification that may limit your ability to substitute w/o re-testing
    2. Sometimes generic jelly bean parts are not as replaceable across manufacturers as you would like and maybe expect. I ran across that with a customer's design where replacing a generic 7400 series logic part with the same from another manufacturer caused the circuit to fail. A case of unknowingly working outside the generic operational guarantees.

    Robert

  • Robert Adsett said:
    that may limit your ability to substitute w/o re-testing

    Just to be clear, that's re-certification testing. 

    Robert

  • Ok, ok already. (re: nit-picking my "protect only the irreplaceable")

    Far and away - "jelly bean" parts have long been "substituted." To avoid the (plight) Robert raises - during design it proves wise to "bring in" multiple such parts - from several vendors - and confirm (early) their "substitution" capability.
  • Once bitten, twice shy.

    A side note since are talking about certification. These units have SW so once you change that you likely need to re-certify in any case. There's a good chance any unit with a micro needs periodic re-certification.

    Robert