Friday, September 06, 2019

Migrant Worker - Periodic Tech Report

Some of you have seen this already - on the Skype feed or splattered throughout my random e-mails. Forgive me for the duplication if you are one of those folks.

Yes, this needs more work but even in this early form it is relevant so I thought it was worth a share. Lately I have been blabbing a lot about the wideband gap semi market(s) and the potential ahead. Some evenings, before I go to bed, my brain starts randomly churning the math that is driving this market. I never really get to the conclusion. Time to build and ramp wafer production line (including reactor), time to design and qualify device, time to make boule, time to prepare wafer, time to process wafer, test time, package time... final sale time - whch depends on customer. On and on and on....

All this number crunching/grinding to arrive at something so basic - # of wafers needed to produce X number of useable devices for X number of use cases. That's oversimplifying things because there are more X's in the actual exercise.... Those that are in this market must be seeing something because they are talking about 30x capacity growth over the next five years.

Sometime, perhaps in the coming days, I'll share a bit of my ciphering with you. Right now it's still a work in progress.

In the meantime, here are a few items worth noting:

ST Microelectronics (STM) made some brief comments about SIC (silicon carbide) at Citi's Technology Conference (so far, I do not see a transcript - but it will probably get posted). Here's a link to the audio:

First, I want to say it's hard (at least for me) to understand,. My ear for the French accent is not that great and he is speaking in English!

So you don't have to listen to it all, allow me to put it in a nutshell.... (the SIC part starts @ around the 26 min mark):

For STM the key customer, right now, is Tesla (TSLA) -- " $100m last year to $200m this year. The longer term goal - (this has to be total marketshare?) is $1 billion by 2025 out of a total market of $3 billion. (If this is total market for SiC devices I think this low, way low - but, what do I know?)

Many of you know how I feel about Tesla - the situation is tenuous. That said, the overall five year and and even longer trend away from ICE (internal combustion engines) is clear. I admit that Tesla kicked this all off - no doubt about that. Whether or not they remain the ultimate winner or remain the biggest buyer of SiC components remains to be seen. I have my doubts.

For the moment the growth in STM's SiC business is mostly Tesla but they are now in production with SiC MOSFETs and diodes for 9 customers. Today 33 projects are ordered which will start in production in the next 3 years (18 automotive, 15 industrial).

STM expects to complete the acquisition of Norstel (SiC wafer maker) this year. They are planning to start internal SiC wafer production in 2021. Side note, which you might probably be aware of - STM has a wafer agreement with CREE (who doesn't?)....

STM was supply constrained last year and couldnt even sample to the industrial market. That has loosened up and industrial has a "big appetite" for SiC.

There might be more - will go over again.


I also went through a few other Citi Tech transcripts this AM - Intel (INTC), Applied Materials (AMAT) to name but two.... These two, along with many others, are saying (hoping?) that the industry is bouncing along a bottom and the future is starting to look more promising. This, of course, is a good thing but it needs to be understood that saying it is happening and it actually happening are two different things. Keep in mind that projecting optimism is part of the nature of the technology beast.

I must admit that I am a bit more optimistic about the future but I question whether or not I can time the turn. Personally, I think it's early and lean toward the idea that it really starts in the middle/latter part of the 2nd quarter next year - 5G rollout in higher gear for sure - memory market, perhaps, bottomed out, semi equipment, maybe a bit longer to recover, but no longer crashing lower.

Calling a bottom at this juncture is, to me, speculation - caveat emptor.

Speaking of bottom callers - Microchip (MCHP), another player in the SiC device market, reaffirmed their "slightly up" guidance for this quarter. I can't recall the last time they did this - the annoucement did not really say much other than they reaffirmed what they originally said. Reassuring.... hmm.... Some of this pickup has to be seasonal. Whether or not it turns into a resilient bottom remains to be seen.

Applied Materials (semi-equipment) pretty much said things were stabilizing during their Citi presentation. No mention of SiC but I need to go through that one again. Applied does talk about SiC on their website: Notable... yes, indeed. Just the fact that they are mentioning if is significant because small markets cannot - no matter how fast they are growing - cannot move the needle at Applied.

INTC said that this year's slowdown in Datacenter spend - which is really not a surprise if you realize that '18 Data Center spend was pretty big - is bottoming. Enterprise is still weak. There does appear to be some life in the PC market - that's probably due to MSFT ending support for Win7 in January. An actual upgrade cycle? Who'd a thunk?

What I found notable - and this has been said many times - is what Intel is saying. This is my take - see the picture, slice the picture up into puzzle pieces, build puzzle pieces and fit together to reconstruct picture. Okay, that is a weak analogy, I know.

We are in the 3rd wave of the internet/connected tech era. First was PC/Internet. Second was mobile/smartphone/Internet. Today it is IoE or IoT or whatever acronym you choose. From sensor input to and through network to compute analysis (AI is the thing today) and back to network to sensor to useable output.

Yes, there is stuff I didn't put in but I am sure you get the drift. I mean, you can even get into the mintuae of this - the transistor level if you want to get really detailed. Such are systems.

The 3rd Wave - Steve Case is responsible for this moniker. Not that it matters but I tagged it as the Sensorization Wave.

i actually wrote about this about 10 years ago. My essay/babble referenced changes that happened in the ITRS (International Technology Roadmap for Semiconductors) - the roadmap became focused on "systems" vs. process tech. This is nothing special but I remember it vividly. Sadly, but par for the course, I received no plaque, noteriety/visibilty or even the slightest hint of recognition. Even though I am probably the only one that remembers this I still like to pat myself on the back for it. Shameless... I know...

Everybody needs a least one Fan.


The linked article (see below) references a KeySight (KEYS) chart showing the application map for GaN and SiC - it makes sense.... Although, i think things are going to move faster than they expect. Of course, I could be wrong here and the world could end tomorrow - all bets are off if that happens.

KeySight, as you might be aware, is the test and measurement spinout from Hewlett Packard - once known as Agilent. Followers and I - yes,I have a couple - were discussing the outlook for KEYS this past week. KEYS has been on a tear and that will likely continue given their presence/exposure to expanding networks and 5G. More on this on another day.

Anyway, scroll down this page to the second image for the KEYS application map:

Interesting factoid: TTI is a wholly owned subsidiary of Berkshire Hathaway.

Was looking for an article about systems I thought I published on Forbes moons ago - wasn't there but I did find this one and it still cracks meup:

Good grief the Forbes website is a nightmare. It's as if the web developer woke up one day and decided that it would be a good idea to slam the webpage visitor with every available tracking cookie, widget, popup e-mail solicitation and ad module. Unreal...


I only have $0.02 today but I am sharing it. Dont spend it all in one place.

Feel free to visit the Sponsor page if you would like to Support the Mission. Your Support is welcome and appreciated!

No comments:

Post a Comment