You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!




Results 1 to 11 of 11

Thread: ?s on TSMC Conference Call

  1. #1
    Expert
    Join Date
    Apr 2013
    Location
    East SF Bay Area
    Posts
    1,549
    Thumbs Up
    Received: 450
    Given: 425

    ?s on TSMC Conference Call

    Having just listened to C C Wei on the TSMC conference call, I have some questions I hope the forum can answer or elaborate on. How much of a lead do you feel TSM has with advanced packaging using COWOS, InFo, and SOIC packaging methods and how much of an edge is this in the market? C C put some particular emphasis on the SOIC packaging technology. Also the other aspect I found surprising was MEMs/sensors at 22nm he specifically mentioned. I was under the understanding MEMS were specifically at larger nodes. Any clarification of MEMs and their uses at this node and their role would be appreciated. Thanks

    Addition: Mems are the future and if TSM becomes the leader in high end MEMs, their partnership with Apple in medical will open a market of staggering size(7 trillion dollars) of which they could take 5% easily. This could add 350 billion in market cap to the partnership they split between them. If all other services and technologies the two could offer in medical are taken into account, this figure could be doubled.

    1 Not allowed!
    Last edited by Arthur Hanson; 4 Weeks Ago at 06:53 PM.
     

  2. #2
    Top Influencer
    Join Date
    Jul 2012
    Posts
    440
    Thumbs Up
    Received: 263
    Given: 2
    Quote Originally Posted by Arthur Hanson View Post
    Having just listened to C C Wei on the TSMC conference call, I have some questions I hope the forum can answer or elaborate on. How much of a lead do you feel TSM has with advanced packaging using COWOS, InFo, and SOIC packaging methods and how much of an edge is this in the market? C C put some particular emphasis on the SOIC packaging technology. Also the other aspect I found surprising was MEMs/sensors at 22nm he specifically mentioned. I was under the understanding MEMS were specifically at larger nodes. Any clarification of MEMs and their uses at this node and their role would be appreciated. Thanks

    Addition: Mems are the future and if TSM becomes the leader in high end MEMs, their partnership with Apple in medical will open a market of staggering size(7 trillion dollars) of which they could take 5% easily. This could add 350 billion in market cap to the partnership they split between them. If all other services and technologies the two could offer in medical are taken into account, this figure could be doubled.
    Are you seriously saying that the medical market *for ICs* is worth 7 trillion dollars? Equipment/products maybe, but ICs form a small part of this value and MEMs a small part of the IC value, so the numbers for TSMC are nowhere near yours.

    0 Not allowed!
     

  3. #3
    Expert
    Join Date
    Apr 2013
    Location
    East SF Bay Area
    Posts
    1,549
    Thumbs Up
    Received: 450
    Given: 425
    No, the entire world medical market is 7 trillion. Much of medical is still backwards and sorely in need of reform. A surgeon recently informed me you could learn 80% of what he does in six months. Like many fields, medical is protected from progress and cost control by huge barriers. I have had this thesis backed up by top engineers that work in the field. Automation and modern methods could cut costs and improve care dramatically. Also a risk based model has to applied, just like auto insurance.

    0 Not allowed!
     

  4. #4
    Top Influencer
    Join Date
    Jul 2012
    Posts
    440
    Thumbs Up
    Received: 263
    Given: 2
    Quote Originally Posted by Arthur Hanson View Post
    No, the entire world medical market is 7 trillion. Much of medical is still backwards and sorely in need of reform. A surgeon recently informed me you could learn 80% of what he does in six months. Like many fields, medical is protected from progress and cost control by huge barriers. I have had this thesis backed up by top engineers that work in the field. Automation and modern methods could cut costs and improve care dramatically. Also a risk based model has to applied, just like auto insurance.
    So your statement that "This could add 350 billion in market cap to the partnership..." is way off the mark, as I said.

    In medical equipment, because of there being so many safety-related and qualification costs the total value of ICs is only a few percent of the total market value. And regardless of how important you think MEMs will be in the future, their value will be dwarfed by the IC value of computation and communications chips -- again, MEMS will probably be a few percent of total IC value.

    Multiply these together and the total MEMs value is probably something like 0.1% of the total medical market value, which is $7 billion -- still huge, but 50x smaller than the figure you gave, and a lot smaller than several existing IC markets like CPUs, memory, smartphone.

    So MEMs is hardly going to transform TSMCs business in the way you suggested -- lucrative yes, revolutionary no.

    0 Not allowed!
     

  5. #5
    Expert
    Join Date
    Apr 2013
    Location
    East SF Bay Area
    Posts
    1,549
    Thumbs Up
    Received: 450
    Given: 425
    In the future mems robots could do much of the surgery. This is a subject I will cover later and I have covered with a top scientist. These would not be cheap, but better than the current medical system. In the near future we will be bionic people. A surgical mems robot costing 50K plus would be cheap compared to what we have now and better than the Intuitive Surgical robots, which are not self directed.

    0 Not allowed!
    Last edited by Arthur Hanson; 4 Weeks Ago at 06:40 AM.
     

  6. #6
    Top Influencer
    Join Date
    Jul 2012
    Posts
    440
    Thumbs Up
    Received: 263
    Given: 2
    Quote Originally Posted by Arthur Hanson View Post
    In the future mems robots could do much of the surgery. This is a subject I will cover later and I have covered with a top scientist. These would not be cheap, but better than the current medical system. In the near future we will be bionic people. A surgical mems robot costing 50K plus would be cheap compared to what we have now and better than the Intuitive Surgical robots, which are not self directed.
    Maybe, but that $50k robot would have maybe $50 worth of MEMs inside -- that's the reality of how equipment BOMs work. The market value is largely in the end product, not the components, and especially not a small subsection of the components. And the total world spend on medical won't go up that much when it's already getting on for 10% of GNP across first-world countries -- or closer to 15% in the USA (world champions!) due to the terrible cost/funding model there.

    So you can only make a lot of money in these markets by doing something which is much better (or safer, or cheaper?) than an existing solution, which would drag money from existing inferior solutions into your pocket -- it's not a market where the overall value can grow by a huge percentage, unless we want the entire world GNP to be spent on medical...

    0 Not allowed!
     

  7. #7
    Expert
    Join Date
    Apr 2013
    Location
    East SF Bay Area
    Posts
    1,549
    Thumbs Up
    Received: 450
    Given: 425
    I highly doubt a state of the art micro surgical robot would only have a fifty dollar BOM. It would involve many very sophisticated chips and mems using several of the very most sophisticated packaging techniques and only after expending a large research budget and approval process. It would be remotely powered and guided(I've checked this out already with top talent). I would be surprised if it was under 10K. These robots and implants would not be cheap, but a fraction of the cost of current solutions. Also, implants using specialized mems, chips and packaging are not like building an iPhone.

    0 Not allowed!
     

  8. #8
    Top Influencer
    Join Date
    Jul 2012
    Posts
    440
    Thumbs Up
    Received: 263
    Given: 2
    Quote Originally Posted by Arthur Hanson View Post
    I highly doubt a state of the art micro surgical robot would only have a fifty dollar BOM. It would involve many very sophisticated chips and mems using several of the very most sophisticated packaging techniques and only after expending a large research budget and approval process. It would be remotely powered and guided(I've checked this out already with top talent). I would be surprised if it was under 10K. These robots and implants would not be cheap, but a fraction of the cost of current solutions. Also, implants using specialized mems, chips and packaging are not like building an iPhone.
    You're the one who said the whole robot would cost $50k, not me -- and based on this I said the BOM for MEMs might be about $50.

    If the $50k/robot was true then the cost of the semiconductors inside it would certainly not be more than a few percent of the total -- go and look at a BOM analysis (vs. selling price) of any sophisticated instrument if you don't believe me. And most of that semiconductor cost will be on providing the processing/intelligence/analysis (which needs a lot of silicon) not MEMs, which are likely to be a few percent of the total semiconductor BOM. Put all these together (say 3% of equipment selling price is semiconductors, and 3% of semiconductors are MEMs) and you get $50 for the MEMs value (0.1%) in a $50k robot. Even putting both these numbers up to 10% -- which is pretty optimistic -- and the MEMs value is 1% of the equipment cost, or $500.

    So even in this optimistic case, the value to TSMC of MEMs is 1% of the TAM for such robots. If the robots cost more the MEMs are worth more, but the percentage doesn't go up. If the total world market for medical is $7 trillion, what percentage of this do you think might be spent on such surgical robots, not forgetting all the other things clamouring for money (CAT scanners, DNA analysis, customised medicine synthesis, doctors, nurses...)? Let's be really optimistic and say that the robots get 10% of the entire medical budget, or $700 billion.

    Divide this result by 100 and you get an optimistic number for MEMs value in this market -- which happens to be $7 billion again...

    0 Not allowed!
    Last edited by IanD; 4 Weeks Ago at 08:32 AM.
     

  9. #9
    Expert
    Join Date
    Apr 2013
    Location
    East SF Bay Area
    Posts
    1,549
    Thumbs Up
    Received: 450
    Given: 425
    IanD, I really appreciate your input. In the case of the surgical micro robot, I think TSM would produce the whole device on a contract basis since the combination of mems, semis, sensors, power sources and communications would be one complex assembly using a variety of sophisticated packaging processes of which TSM has mastered. TSM is one of the few companies that I feel could make the whole device. If you have any other thoughts, I would appreciate them. I deeply value the input of the SemiWiki community and I deeply value the input the members have given me. Never hesitate to be my devil's advocate, it could save me valuable time and resources. Thanks Again

    0 Not allowed!
     

  10. #10
    Top Influencer
    Join Date
    Jul 2012
    Posts
    440
    Thumbs Up
    Received: 263
    Given: 2
    Quote Originally Posted by Arthur Hanson View Post
    IanD, I really appreciate your input. In the case of the surgical micro robot, I think TSM would produce the whole device on a contract basis since the combination of mems, semis, sensors, power sources and communications would be one complex assembly using a variety of sophisticated packaging processes of which TSM has mastered. TSM is one of the few companies that I feel could make the whole device. If you have any other thoughts, I would appreciate them. I deeply value the input of the SemiWiki community and I deeply value the input the members have given me. Never hesitate to be my devil's advocate, it could save me valuable time and resources. Thanks Again
    If TSMC can "move up the value chain" (uurgh, I *hate* that phrase) then that's a completely different market with a higher value attached. However I would expect that the design of such a module (and software, and marketing) would be done by one of their customers with TSMC manufacturing and assembling the module but final product assembly and testing being done elsewhere, which then moves revenue away from TSMC.

    The total semiconductor value in such a product (even if TSMC grabbed it all) rarely exceeds 10% of the final product price by the time all margins and costs have been accounted for, which could give a TAM of $70B if they got 100% market share (and the robots took 10% of the worldwide medical budget). That's a lot more attractive, but based on a lot of very optimistic assumptions so unlikely to be reached for a long time, if at all. TSMC revenue is about $40B a year right now, so such a product could certainly give them a good chunk of revenue, but probably not enough to transform the company.

    Bear in mind that developing such technology and building such a micro-robot and getting it into initial trials would take several years, then a few more years trialling and testing to get it approved as safe and effective, then several more years to ramp up production and get it into the market in a big way -- in other words maybe 10 years to get to the kind of revenues we're talking about.

    You've got to have deep pockets and a long-term view of success to bet on such a product. I agree that if anyone can do the silicon side of it TSMC can, but whether even they'd be willing to invest in something that wouldn't pay off big-time for 10 years is an interesting question...

    0 Not allowed!
     

  11. #11
    Expert
    Join Date
    Apr 2013
    Location
    East SF Bay Area
    Posts
    1,549
    Thumbs Up
    Received: 450
    Given: 425
    IanD, I feel TSM will evolve into a complete nanotech company that makes a full range of labs, robotic parts, complete mems assemblies having a broad range of uses in many industries since they are the ultimate nanotech company on the planet and this would be just a logical extension that would spread their costs over a much broader range of uses giving them an even larger competitive advantage over their full product line. In this world you are either growing or dying because of the ever faster pace of change. TSM has always grows and evolves and shows no signs of stopping. They do have failures like anyone else in lighting and solar.

    0 Not allowed!
     

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •