AKP AGM
As I expected a significant revelation to be made I flew over from Christchurch for my 1st AGM. As expected it was indeed very interesting, even though the “revelations” were not all that I had hoped I might hear.
I was limited to only being able to
read HC on my iPhone because I had forgotten my login password! My comment below was written just after AGM and in response to other posts. I will have to post it when I get home.
I concur with all of VB’s observations here. These were my own observations.
▪Modified MEMS wafers are currently in the company’s hands from 2 separate suppliers.
- TowerJazz is clearly AP’s preferred primary MEMS vendor for very understandable reasons, and will continue to be a close relationship with AP into the future.
- I did not hear any indication as to who the 2nd vendor was (I see others picked that it was still SONY. The depth graphs are from SONT and TJ with dates from Jan-Apr 2017. Well spotted guys!
- I still do not know who the 3rd party FAB was, but they have the best ability to clean out the cavity under the MEMS components, so the nearly completed wafers get sent there for etching and cleaning as a final step.
I read the interplay between Fred and Danny as reflecting Fred’s enthusiasm and Danny’s cautious optimism. Danny knows that “it ain’t over until the fat lady sings” because he is responsible for the tech in detail and does not want to make claims if he does not have absolute facts to back himself up with. Fred knows that the chips are almost cooked, and the fat lady is clearing her throat, and also that the market needs to know that that is where we are at. Both of them are right, and neither contradicted the other.
Both indicated that finished - and demonstrable - chips were due to be produced very, very soon; “imminent” was the term used by both Fred and also later by Danny I believe.
From here the following milestones can be expected:
- Testbed measurements need to be done next to be sure that all of the steps have worked as expected.
Danny made the point that both they and the MEMS fabs were “certain” that the wafers were up to spec. This was a different question to whether the designed-in specs produced exactly the behaviour that was required to deliver the expected and designed-in performance. This could and would only be known after both testing and final integration. Hence Danny’s comments about whether the final steps would show some unexpected and unanticipated problems. This will be the very first time in the history of the world that what they have designed will be shown to have been done in exactly the right way - or not. However, I think it is clear from all the technological innovation and the testing done to date that we can be sure that we are >99% of the way to proof positive. I think it is very reasonable to expect that the designers who have brought the tech this far are careful enough and capable enough to pull this off. Does that leave any room for uncertainty? Of course it does. I got on a plane to get here. Was there any chance anything could have gone wrong? Of course there was, with every plane that flies, but that did not stop me from flying. (Getting me home saw a delay of over 2 hours when my plane broke down in Fiji and had to be replaced by another in a hurry! But I still got on and flew home in it.) I would put this into the same context as other cutting edge science - something I am regularly involved with. Is this a moon shot? Well, no, I think its actually way surer than a moonshot. It’s just hard and painstaking work, that requires every t being crossed and every i being dotted.
- After measurement/testing the chips go to be integrated and packaged.
- The whole integration development process was expected to take 4-8 weeks the first time through, so this should be a much quicker process.
- I am picking that wafer testing will take maybe 2-3 weeks, and integration/packaging maybe another 2-3 weeks.
- After receipt of final chips I expect some time will need to be spent making absolutely sure that chips will be thoroughly evaluated and driven to destruction to ensure that performance and reliability etc are as anticipated. Maybe another 2 weeks?
At the AGM I queried the possibility that by making the membrane thinner so it would be easier to move electrostatically with a lower voltage this may also have made it more delicate, and hence less durable. Danny’s reply to my question in the meeting, as well as afterwards, was that this change was still well within the design-envelope for the physical scale of the device. I found this quite reassuring.
- I am picking therefore 6-12 weeks. From the emphasis on “imminent” provided by Fred and Danny I would not be surprised if the next announcement comes at the lower end of this estimate. “Imminent” suggests something way shorter than AP has ever indicated previously.
- After this Danny will presumably be back in Sydney to a special meeting to show us the chips, and hopefully to demonstrate some of their capabilities in elementary configurations and commercial device mockups - at least that is what I’m hoping to see and hear. Especially if I have to fly over again for the show!
▪NB: This timing is, I hope, an insightful guess on my part, but nothing more. DYOR.
- The shift in business stage from being a start-up developer of a new tech to being a full-scale supplier of mass produced electronic components will inevitably and necessarily require a change of business model, and this is clearly something AP has already considered and planned for.
- There will be a capital raising after the product has been demonstrated and launched as a product. Just moving through to being a supplier involves a massive scale up of production volume that extends out several years before AP is approaching its maximum potential revenue stream - which is very likely to be substantial, and not without cost, so CR is essential. I just hope we are not overly diluted in the process and get an advantageous shot at it too.
- I thought Danny clearly indicated that AP intends to remain a FABless vendor for the foreseeable future. Setting up a FAB costs billions. AP is not in that league and never intends to be. Remaining FABless allows it to remain nimble and responsive and to spend money prudently and efficiently. To be a preferred supplier to most large market customers means that AP would have to have second sourcing for all processes to ensure guaranteed continuity of supply. It would thus need to still contract work to at least one other FAB.
I suspect that Share Price movements leading up to and following the AGM seem to mostly reflect reduced expectation of a feeding frenzy that would have followed a more definitive announcement. In that light some have got out quick, others have jumped in quick, so we are still in a holding pattern.
We live in interesting times, for which patience is still the order of the day!
Here's a couple of images of the chips FYI.
The chip in my hand is completed but unlabelled.
The chip at the top of the box has been poked by a curious finger and damaged.
The one at the bottom shows the integrated but unpackaged chip.
View attachment 601725
View attachment 601728
(20min delay)