In 1968 two unhappy engineers in California bailed on their jobs to try their hand at something new.
One, Robert Noyce, earned the sobriquet “the mayor of Silicon Valley” and the other, Gordon Moore, named a physical law after himself.
They founded Intel and were among the first ‘nerds’ to make a fortune from personal computers.
Down in Australia at the time, inventions were more hands on, including the blow up cask for cask wine, the ultrasound, and the inflatable slide for escaping aircraft.
But the country missed the semiconductor boat, says Australia’s top quantum computing boffin Dr Michelle Simmons.
In evidence of this, only four of the 11 companies who say they have a hand in the lucrative semiconductor trade are Australian.
Almost all of those are working with early stage tech — the kind that isn’t going to make money for some years yet — but they all believe they will be the next big thing in tech.
Most of these companies were in the same room-with-a-view in a Sydney office tower on Thursday last week, pitching to the uninitiated in an event pulled together by Pitt Street Research director Marc Kennis.
A who’s who or a sum total?
Of the semiconductor companies on the ASX, four are true-blue Aussies. Two are Israeli and five are blow-ins from the US.
Two sell equipment for making semiconductors, in a play on the picks and shovels in the goldfields theory.
And three actually deal in MEMs, or micro-electro-mechanical systems. These are chips which are made of miniaturised moving parts such as accelerometers, pumps, and force sensors.
Victoria-based Bluechiip makes MEMS chips for tracking very low temperature biological samples, down to -196 degrees Celsius. The chip is embedded into a test tube which is then sold with a hand held reader.
The company has one customer, a three-year, $16m supply deal with US lab plastics manufacturer Labcon.
Managing director Andrew McLellan says they compete with handwritten labels and barcodes, although Internet of Things companies like CCP Technologies (ASX:CT1) are trying to compete at the slightly warmer — -80 degrees — level as well.
- Subscribe to our daily newsletter
- Join our small cap Facebook group
- Follow us on Facebook or Twitter
Their chips cost $1 each and freezers can contain tens of thousands of samples, but the next generation of the tech will provide electronic tracking of temperature changes and a blood bag maker has already said they’ll pay $10 a chip for that.
“We are on the road map of all the majors, I know that for a fact.”
They met with one in Shanghai last week, in fact.
Bluechiip is one of the few semiconductor stocks making money, pulling in $271,000 in revenue in the March quarter.
Another Australian is BluGlass, which managing director Giles Bourne is pitching as a licensing play to manufacturers
They’re developing a new semiconductor manufacturing tech using a process called remote plasma chemical vapor deposition (RPCVD), which they say is cleaner and cheaper, and because it’s done at low temperatures most useful for making LEDs.
Bourne says most of the work is research right now, but they do make a small amount of money ($152,000 in receipts in the March quarter) from contract manufacturing.
But he also says they’re “actively engaged or talking to several” of the top 10 LED makers in the world.
They’re “well positioned” for commercialisation but there’s no word yet on when that will happen.
K2 Energy (ASX:KTE)
Australian IP investor K2 Energy is dabbling in semiconductors via a shareholding in NASDAQ-listed Atomera.
K2 bought the exclusive worldwide rights to the Mears Silicon Technology (MST), a new tool which can be used to make semiconductors, for all solar energy applications by funding an R&D program.
As an investor, they aren’t making any money from it yet, but say a Japanese and a South Korean company have signed licensing agreements for the tech.
Silex Systems (ASX:SLX)
Silex is an R&D company whose primary asset is a laser uranium enrichment technology.
But through a subsidiary called Translucent they’ve licensed a set of semiconductor materials known as crystalline rare earth oxides (cREO) to UK company IQE to turn into chips.
They did the deal in March 2018 and was paid $US5m in IQE shares and also get a perpetual royalty of 3-6 per cent on the sale of any products that use cREO tech. Minimum annual royalties start at $US400,000 in fiscal year 2020.
Audio Pixels (ASX:AKP)
The Israeli directional sound MEMs chip maker has been persevering with its difficult-to-produce speaker for about 13 years.
They’ve raised over $40m from the market and have a market cap that usually hovers around $500m.
Led by chairman Fred Bart, a former champion poker player, Audio Pixels has for years been developing a chip for a new kind of digital speaker that can deliver boom-box sound in a mobile phone, that can be channeled in a just a single direction.
“Over the last 10 years we are the best performing stock on the ASX,” Bart said at the semiconductor love-in last week.
The company has run into a series of delays over the last two years in manufacturing a viable prototype for testing.
Thee company was not able to say when it would have a product to commercialise, only that they are “very close”.
Weebit Nano (ASX:WBT)
Fellow Israeli is Weebit Nano, one of two companies on the ASX working with ‘ReRAM’ technology which is a new middle way between DRAM (dynamic random-access memory) — expensive but fast — and NAND Flash — cheap but long memories.
Pitt Street Research’s Marc Kennis says as the limits of Moore’s Law are reached — the one named after Intel’s Gordon which observes that the number of transistors in an integrated circuit doubles about every two years — new storage options need to be found.
He says ReRAM, or resistive random-access memory, means the whole surface of a chip can store memory cells and thus speed up computing while using a very low current — making it cooler too.
Weebit signed a development partnership with chip designer Silvaco in January and expects to be making money from their chip come 2021.
4DS Memory (ASX:4DS)
Opening the US line-up is 4DS Memory, the other company working with ReRAM.
Director David MacAuliffe is the Australian representative for the company and, coming from a biotech background, is keen for investors to view 4DS as a proof of concept development play which will eventually be acquired.
They have partnered with global hard-drive manufacturer Western Digital as a research partner, who McAuliffe says would also “be our ideal exit partner”.
He didn’t imagine that Western Digital would push them aside — not after six years of due diligence — but they are 4DS’ only potential acquirer so far.
4DS plans to have analysis results from its preferred chip, made by imec, ready in June.
Like a biotech they are in R&D phase, meaning they’re not making any revenue, and as they’re not planning to commercialise their technology themselves, a payout for investors depends on if and when a bigger company chooses to acquire them.
Pivotal Systems (ASX:PVS)
Pivotal only listed in July 2018 and actually makes gas flow monitoring and control technology used while making semiconductors.
Semiconductors require precise flows of gas to shape the underlying wafer on which the circuits are built, and Pivotal supplies that equipment.
In the March quarter they took $US3.2m in cash and were left with a cash balance of $US13.5m and no debt. It’s spending hard — $US6.6m — on building up its manufacturing base.
It expects revenue for the half year to June 30 to come in at $US8-9m.
Like Pivotal, Revasum sells equipment used in making semiconductors, a selection of very expensive automatic machines involved in grinding and polishing.
The Silicon Valley company listed in Australia in December because they only wanted to list as $30m company, a size they’d not be able to do easily in the US.
Revasum is also making money, taking $US8.7m in cash in the March quarter.
However, in April they also issued their first warning: half year revenue is likely to be $US15m to $US16.5m instead of the $US20.5m promised in the prospectus, because of delivery delays.
Brainchip is designing chips that don’t need server backup to keep learning, which has obvious uses in driverless cars or IOT devices that need to use less power, such as those operating in the outback where regular battery changes are unlikely.
The company is making a small amount of money as it commercialises its design, but has run into problems with shareholders upset about executive remuneration.
Ten directors and executives were paid $1.9m in salaries and fees, while share-based payments hit a total of $5.9m.
In April the company outlined its response to shareholder concerns, saying CEO Lou DiNardo would take options instead of performance rights and management salaries would be reduced, although not by how much.
Sensera made it name with MEMs cow ear tag trackers, but has expanded into software and chips in the health sector.
The MEMs chips are for blood monitors and heart pumps which are inserted inside the body.
CEO Ralph Schmitt says they’re starting to see revenue from the heart pump grow from now, and their ‘organ on a chip’ should start making money from 2021.
Schmitt is hoping the accompanying software-as-a-service division will start kicking in 15-20 per cent of revenue in from FY2020.
Sensera made “record” revenue of $US3.5m in the March quarter but did have to take out a $US1m loan and raise $3m from investors in order to keep things ticking over.
But the future is quantum
The farther off opportunity, but one likely to breed more Australian business, is quantum.
Simmons, last year’s Australian of the Year, says while Australia missed the semiconductor industry, it is at the forefront of quantum computing.
“Australia is incredibly strong in quantum,” she said during a semiconductor company showcase on Thursday last week.
“It literally is a whole new industry.”
An industry of researchers is growing up in Australia which, if it follows semiconductor history in the US, should form the basis for a commercial industry in this country in about a decade.
And a few months ago CSIRO chief scientist Dr Cathy Foley started work on a national strategy for quantum computing.
Quantum computing is where instead of building a circuit using tiny parts, you use atoms. Without going too deep into the mechanics, if a classical computer can work through code one by one, the extra power of a quantum circuit does all at once.
Simmons says the part of the reason behind developing quantum computing for commercial applications is that Moore’s Law is reaching its upper limits. New ways of making computing faster need to be found.
She says any industry that uses data will be affected, but it’ll be a decade away.
She bucketed these into four areas:
- optimisation, such as couriers looking for the best route;
- machine learning such as in electric vehicles learning new road conditions and driving styles;
- quantum simulation where people like Simmons use it to create more complex systems;
- and prime number factorisation, or code-breaking as the lay-people in the room would understand it.
Trading as we know it will change as quantum computers will be able to find the perfect return for the perfect level of risk in seconds.
As yet, quantum computing is such a new niche it makes up a small part of investable companies such as Alphabet, IBM and Lockheed Martin, and investors will need to wait to see if leading quantum hardware startups such as QCI, Rigetti or IonQ will survive to IPO.