Eulogy for semiconductors in Canada

The last two nails were hammered into the semiconductor business in Canada recently with the acquisition of GENNUM  by Semtech and ZARLINK by Microsemi (both out of California). This follows the recent acqusition of SiGe Semiconductor by Skyworks and earlier TUNDRA by IDT. For all practical purposes we are left with no significant size Canada-based and Canadian controlled semiconductor company in the country – back to chopping woods, digging minerals and pumping oil (not that there is anything wrong with that…)
Maple leaves falling
Oh, sure, there are still some left-overs and a few pocket-size and interesting niche-y microelectronics firms but nothing of the scale that would make a difference at the national level. Does it matter? History will tell but it sure is a bit sad for those of us who were part of the ride and, well, after all, who really wants his future to be just about running a branch-plant?

Thus, appropriately here is a nostalgic look back and a brief crash course in the history of the semiconductor industry in Canada.

The beginnings of it go back to the early ’70s and the story of MicroSystems International Limited (MIL), initially a government inspired venture with Northern Telecom (Nortel). When MIL was winding up in 1974 it spun out two important seeds: Semiconductor Components Group (SCG) as part of Nortel/BNR and a pair of two budding entrepreneurs Terry Mathews and Mike Cowpland. They started MITEL whose semiconductor division (Mitel Semiconductor) later on became ZARLINK. At about the same time Wally Pieczonka and Doug Barber started Linear Technology Inc (LTI) in Burlington, Ontario which was renamed GENNUM later on. These three companies became pillars of the foundation on which pretty much all the rest (with an exception of PMC-Sierra) of Canadian semiconductor industry was built.

The golden age lasted about two decades (1980-2000) with the ’80s being particularly heady days as the industry was young and rapidly growing. When I joined Nortel’s SCG in the early ’80s as a member of its R&D team, Nortel had a vertically integrated semiconductor operation. It involved not just the chip design but also manufacturing of silicon, process and device technology, packaging, testing and design automation. The synergies in such an environment were just immense. We were doing world-class engineering and, most of the time, money was no object. As a young manager I was nevertheless able not just to collaborate but also sponsor leading-edge research at Stanford University, Carnegie Mellon and a number of Canadian schools.

In this fertile environment a number of world-class inventions and products were developed such as the CCD imagers technology by Jim White and Joe Ellul, some of the industry-first CAD tools such as AUTOLAY and SYNFUL by Stan Jedrysiak, as well as a number of first complex telecom chips that powered the digital world. Some memorable management figures from that time were Lloyd Taylor, Graham Sadler, Geoff Shrank, Dave Lawrence, Adam Chowaniec and later on Ken Bradley and Claudine Simson. Similarly Mitel Semiconductor (ZARLINK) built its own semiconductor fab in Bromont, Que and moved into the merchant semiconductor business led by Doug Smeaton, David Brown and later on Kirk Mandy. At the same time Linear Technology (GENNUM) led by Wally Pieczonka achieved a dominant world market share of 65% as a supplier of hearing aids chips.

The spillover effect of this critical mass on the broader Canadian scene was quite substantial as it fed a large thriving supply chain, stimulated world-class research at the universities including funding university Chairs (Carleton U) and spun off a number of startups. Some of the better known included Siltronics (Gyles Panther), Mosaid Technologies founded by Dick Foss and Bob Harland, DALSA founded by Savas Chamberlain, CALMOS (later on transformed into Newbridge Microsystems and finally TUNDRA, acquired by IDT) founded by John Roberts, Genesis Microchip founded by Paul Russo, and many others. Among them was ATMOS Corp which I founded in mid-nineties. There is a famous chart, created and supplied by Doyletech, showing the family tree of locally-generated technology companies and pretty much 90% of them can be traced back to NORTEL.

In the ’90s there was an important joint industry-government initiative at the time called Strategic Microelectronics Consortium (SMC) run by John Roberts. It really provided a strong boost to the growth of the fledgling microelectronics industry in Canada. In addition, the Canadian Microelectronics Corporation (CMC – currently run by Ian McWalter), mostly government funded,  was created to stimulate and support electronics research in Canadian universities.

Unfortunately most of it started to slow down around the year 2000. After a good, long 40 year run, the semiconductor industry was maturing (not unlike what happened to the automotive industry). Some players, such as Mitel Semiconductor (ZARLINK), caught the change and transformed themselves from captive suppliers into merchant semiconductor players. Unfortunately, the largest of them, Nortel’s SCG, failed to achieve that, mostly due to lack of management leadership. As a result, this primary engine of semiconductor growth and expertise in the country was sold off to STMicroelectronics and following a classic pattern drifted away never to be seen again!

There is no doubt in my mind that the failure of the leadership of Nortel’s SCG at that time to spin off its microelectronics business as something like “Telecom Semiconductor Inc”, was a main trigger for the subsequent decline and slow disappearance of the semiconductor industry in Canada. We have simply lost critical mass. A number of talented highly specialized engineers and researchers moved away in search of the jobs in California’s Silicon Valley and other places around the world. It’s a pity but that is what lack of foresight, vision, circumspection does to a country…

For a little while though “Times they were a-changing” we got a bit of a second backwind towards the end of the ’90s with the days of mushrooming semi start-ups nourished by the multi-million dollars investments from the burgeoning VC firms. This was the era of a new type of semiconductor company, no longer large capital intensive and vertically integrated, but so called fabless semiconductor company. Among the better known started at that time were: SiGe Semiconductor (founded originally by John Roberts), Skystone (Antoine Paquin & Stefan Opalski), Solidum (Feliks Welfeld), Quake (Dan Trepanier), ATMOS (Paul Slaby), Lumic/Atsana (Luc Lussier), Philsar, Extreme Packets and IceFyre. All of these companies have been acquired since and are mostly gone from the Ottawa scene.

So, what have we got left?
What is the landscape after the battle? What we have is a number of small and medium size firms built up on the remnants of the previous firms. They are mostly profitable and sometimes hugely so. They usually do not forge new product frontiers or aggressive innovation. Instead they tend to provide services and capitalize on the know-how and expertise developed by their predecessors.

The most successful is definitely a cluster of what one could call “Semiconductor IP protection and licensing”. This includes:  MOSAID  (recently renamed “Conversant Intellectual Property Management” was founded by Dick Foss, later on run by George Cwynar and currently by John Lindgren), Wi-LAN (Jim Skippen), UBM TechInsights (previously Semiconductor Insights – a spin-off from MOSAID built-up by Terry Ludlow and Doug Smeaton), Chipworks (Terry Ludlow), Global Intellectual Strategies (GIS – Pierrette Breton). It is interesting to note that all of these companies can be traced back to MOSAID – clearly Dick Foss must have done something right laying down this foundation!

In addition, there is a sprinkle of design services and IP product companies such as TSMC Design Centre (as a result of EMT (Sreedhar Natarajan) acquisition with its origin in ATMOS Corp), Kaben Wireless Silicon (which we have built-up significantly during my recent 3 year CEO run), SiDense (founded by Wlodek Kurjanowicz and run by Xerxes Wania), CogniVue (with its origin in Lumic/Atsana and currently run by Simon Morris). Outside of Ottawa the significant players include PMC-Sierra in Vancouver, DALSA, Fresco Microchip (Lance Greggain) and ViXS (Sally Daub) in the Toronto area. The industry has its representation through the ITAC SMC Council currently coordinated by Iain Scott.

So, now, what does the future hold for semiconductors in Canada?
The business has changed and the glory days appear to be over and are not likely to come back. It is a different world now and no amount of nostalgia is going to change that. In particular, the old business model of fabless semiconductor companies is, for all practical purposes, dead,  when it comes to start-ups and emerging companies (you can find more about it in this presentation: http://www.design-reuse.com/exclusive/kaben/)
There is a need for new approaches that have a chance to bring significant ROI justifying investments.

Just because the old ways of doing business are no longer applicable, this does not mean there is not a need or a demand for semiconductor start-ups and their innovation – quite contrary! But the way we go about it has to be different. To avoid repeating myself, I refer you to this article: http://www.eetimes.com/electronics-news/4074052/Letter-to-the-editor-IP-cars-share-common-ground

PS
Here is a bit of an inside scoop on the Zarlink story:
And here is an info on the GENNUM acquisition:
Advertisements

Comments

  1. Electricity and semiconductors industry getting new updates. Japan is working now Fabric with micro solar cells generate electricity. It is great. TOKO is working on fabric that can actually make wearer to generate electricity. Business Brecorder addedupdates.Asic Design

    Like

  2. Seamus J. Laverty says:

    Dear Paul

    While “googling” for Graham Sadler your blog popped up. Do you know how he is? He was my boss at Northern Electric / BNR from 69 to 72.

    Your blog mentions that the Canadian technology explosion all began back in early 70s. You are right and I may be able to add a little to that.

    With my bride of 5 weeks, I at age 26 arrived in Ottawa (1969) to find an obsolete Al gate process but by using my PhD from Queens University Belfast I got a “pilot line” silicon gate process working in the NE labs by early 70. Switching Division liked the sample circuits I provided and wanted me to get MIL into production mode. I explained to Frank Eedy (VP) and AG Sadler that I didn’t have production know how but Drs Noyce, Moore and Grove had a start up called Intel and they had the know how; in particular Glass Reflow for step coverage and planarization. Up to know those names were but authors of seminal journal papers. Within two months a deal was done for $2.7 Million Canadian and I was to run the 1970 / 71 technology transfer up to MIL.

    I had arrived in Mountain View with my wife and 7 week old daughter in what Dr Noyce called a Grandpa car; a large Dodge Monaco. Dr Noyce made me very welcome and had me to his home. He insisted on my leaving the Dodge and taking his Cougar while he found me a Camaro. Oh, the Streets of San Francisco; shades of Steve McQueen in Bulitt.

    However I didn’t get on too well with Dr Grove; he was suspicious of my occasional “stray” into the Schottky TTL bipolar fab area! but his book “Physics of Semiconductor Devices” was invaluable. Gordon Moore had a fiercely penetrating gaze that fell somewhere between the paternal Noyce and the suspicious Grove. Dr Noyce tried to convince me to invest $50k into stock. I felt it was too risky. Oh dear me!

    Together with layoffs the Canadian money allowed for the 1103 1024RAM and put Intel end of year accounts in the black for the very first time. I ended up with a BNR business card that said Manager MOS Technology. When MIL was eventually folded the Si Gate technology was the seed of several new enterprises

    Twenty years later I found myself advising the Intel’s legal team in London on their defence against the1964 Boyd Watkin and Michael Selser patent on self aligned gates, discovered by GI in its archive from the Philco Ford acquisition. The legal argument centred on: would a practitioner of the art in 1960s have recognised polysilicon as a contender “ refactory metal” and would the same person have been able to transpose “ align” with “register” as meaning the same?. Whatever happened to those two inventors?

    Other BNR projects in the 70s were a legacy wide bandwidth analogue pnpn crosspoint array that in collaboration with Switching we made into a small PABX. Another project was Flip Chip by gold beam-leads onto ceramic substrates. I also had CCDs and imaging work led by Dr Ali Ibrahim

    Unfortunately family circumstances back in troubled Belfast required that all four of us return home and that is where I have been ever since.

    In 76 Graham tried to pry me out of my university nest to come over and do some work on 3-5 semiconductors. In 89 when over visiting the Nortel plant at Newtownabbey he called for a social visit. We had got on great together and had many memorable business trips. So now well into retirement my mind turns to wondering about Graham and all the other great guys of those fun days. Did he ever get his Irish pubs? Oh a funny story. When over in 89 despite the “Troubles” he was forever photographing pubs; unfortunately one was next door to a fortified police station; a snatch squad came out and bagged him!

    Other names you may have heard: off Rudy Kreiger; he sorted out the Na+ problem on gate threshold with a touch of HCl. Dug Coulton, Bob Ferguson, Alan Aitken, Eric Van Tongerloo, Ray Hoare, Keith Richardson, Jeff May, Terry Caves, Dave Vincent, Bill Westwood, Tony Springthorp, Kevin Ford are others from my era.

    I know Terry Matthews and Mike Cowplan got over their false starts and made good, very good in fact. Indeed the last I saw of Alasdair McArthur, one of my process engineers was on BBC TV news with Alasdair showing Charles, Prince of Wales around Terry’s Newbridge plant. So Alasdair must have done well too.

    Alasdair reminds me of another story. Taking an afternoon off from the 71 Washington ED conference to sight see, saw Alasdair and myself heading towards Arlington but a “short cut” through the Pentagon car park eventually had us walking around floor 5 of the Pentagon. We were soon apprehended by two MPs, interrogated by officers as to our mode of entry, camera film confiscated, given coffee and biscuits and eventually delivered back to the Hilton curtsey of US army. One of several mischievous but memorable adventures.

    Seamus J. Laverty

    Like

    • Michael King says:

      Hi Seamus,

      Few people realize that Northern Electric was making semiconductor components in Montréal well before MIL was started. In fact, NE made what were regarded as the best Silicon PNP transistors money could buy (it was all in the passivation). When Northern had the bright idea of exploiting their processing know-how by making ICs in Ottawa, many of the process engineers in Montréal refused to move, and the first wave of Canadian semiconductor process engineers went South.

      Some important people ended up in Motorola in Phoenix, in fact the man who ran the bipolar oxidation/diffusion processes there was an es Northern man (forget his name now, being so long ago).

      Also, there was a very innovative semiconductor R&D group operating out of BNR Lab 1. I was a member of this group from 1969, and it was one of the best working environments I ever experienced. We helped set up some of the processes in the New MIL factory.

      Best regards to all of the old semiconductor gang.

      Mike King

      Like

      • Seamus J. Laverty says:

        Hi Mike,

        Yes I recall your name but the years have deleted what you were up to. You never came under my umbrella. Do you recall any of the names I mention particularly Eric van Tongerloo and the pnpn air isolated crosspoint switch. John Rywack was another name I interfaced with though I recall he was more circuit design and kept pestering me about Si gate process design rules. I was more interested in incrementally reducing gate Vt; he was looking for consistency!

        Yes you are absolutely correct about the pnp bipolar in Montreal. It is largely forgotten about and passivation of course was the trick. prior to NE labs I had spent time with Texas Inst in Bedford England and doubled the TTL yield by gettering the oxides and introducing an oxygen- nitrogen 450 C anneal, All tricks learned during my PhD and subsequently applied at the NE labs prior to BNR in 71. I had to insist on liquid sources and heli-arc welded stainless pipes as bottled gas was not up to par.

        Do you remember Dave McLauren the chief technician ( he had a missing thumb); now he was a star and gave great continuity as staff came and went.

        I arrived August 69 and wrote a report rubbishing the existing Al MOS process as obsolete and recommended a Si gate pilot line. I was made project leader and a budget of around $1m Canadian to get set it up. I remember that day very well as after the meeting Graham and I went for lunch in the canteen and graham said ” I think today we can aford the $1.58 lunch” . That pilot line eventually led to the 1970 Intel deal etc. So you must have been at MIL end of the transfer while I was in Intel. It was a great group.

        Graham regularly referred to the Montreal operation and to a unit making “obsolete devices” as a great money spinner. I think it was Shearer St in Montreal that he had me go to see the rotary dial assembly line. Dials made for 58c. Could a push button array and the associated ICs compete on price.?

        So Mike great to hear from you. and yes they were great days.

        Just heaard from Paul and John that graham passed away last year.

        Seamus

        Like

      • Michael King says:

        Hi Seamus,

        Nice to hear from you.

        Yes, I remember Erik van Tongerloo. We shared adjacent desks at the far end of Lab 1 in the early days. John Rywack was my supervisor at one time.

        I was a BNR employee from its inception in 1972 until my birthday in 1982, when the silicon R&D group, plus the fab line were willy-nilly transfered to Nortel, so that they could use it to make ICs for their equipment. I think there would have been tax issues if they had used parts from an R&D facility – plus they wanted full control.

        We then moved to Corkstown and built MOD2 (using many of the machines that had been mothballed from MIL). Later we were involved in building Lab 5 as an extension of the Corkstown building, and installing MOD3 (1 micron capable).

        Exciting times, and very fulfilling.

        Best regards,

        Mike.

        Like

      • Hello Michael;
        My father, Ian Griffin, worked for Northern Electric/BNR/Nortel from 1969 to 1993, and his time there (mostly in Switching and Standards Coordination) left me with a love of telephony, and fuels my currently collection of NE/NT telephone equipment. I though you’d be interested to know that I recently purchased a double-rack steel cabinet that was originally a Diode Test Station at the MIL lab in Montreal, before it ended up at the Corkstown Road campus. Of course, there was no test equipment in it, but it has been a cool piece of history to acquire for my own surplus Nortel test equipment. I bought it from Wayne Getchell, who mentioned he once worked at MIL. Funny how this stuff goes around and around.

        And with reference to John Tyson’s book mentioned by Paul Slaby, it’s available on Amazon.ca and you can search for it using the ASID of B00HTNMQJO.

        Regards,
        Tim Griffin
        Ottawa, Canada

        Like

      • Michael King says:

        Hi Tim,

        thanks for that,

        It is circular, as you say. I worked with Wayne Getchel on several projects.We traveled many times together to the Boston area to discuss project issues with the chip suppliers.

        We now meet often at the Nepean Sportsplex where Wayne takes an exercise class given by my wife,

        Regards,

        Mike King.

        Like

  3. Thank you very much, Seamus! Your reminiscences are gold for the history of Nortel, luckily mostly the glory days and not the tragic ending…
    By the way, one of my friends, former BNR VP, John Tyson is writing a book, kind of a history of Bell Northern Research. I will pass on your recollections to him – he might be able to include some of it.

    All the best for now and if you ever get to Ottawa, by all means let me know. I would be happy to meet in person.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: