The recent UK Innovation Strategy, published by the UK Government, highlighted semiconductors as a critical technology area for increased investment, with the recognition that by the mid-2030s UK companies could play an increasingly central role in a supply chain of “acute geopolitical importance”. But many are sceptical about this goal, especially when they see the eye-watering cost of leading-edge semiconductor fabrication plants.
Advances in semiconductor technology have had a massive impact with a myriad of electronic devices and are at the heart of recent advances in healthcare, including sophisticated surgical robots and clever Artificial Intelligence (AI) devices.
The silicon chips used in these applications often cost $100 million or more to develop and take years from concept to production. To understand why, we need to consider the underlying technology for making them: the production process starts with an expensive, highly refined wafer of crystalline silicon, and then modifies the material to create the necessary characteristics for semiconductor device building blocks such as transistors, before adding further material layers to interconnect these devices for a complete integrated circuit.
Moore’s Law has driven progressive reductions in minimum “feature size” from 10µm to under 10nm – so now 1000x smaller devices deliver proportionally higher performance and greater circuit complexity - but require increasingly sophisticated and expensive machinery to make them. Over the same timescale, wafer size has increased from 3” to 12” (300mm) diameter and the number of layers in each chip has grown from less than 10 to nearly 100! All of this has taken a fundamentally expensive process and continually increased the cost per unit area – while of course delivering a lot more functionality within that area.
Not only has this led to a massive concentration of supply, with silicon chip manufacturing requiring enormous “mega-fabs” (often subsidised by local governments) in order to achieve the necessary economies of scale, but it has also driven extremely long production cycle times – typically 3 to 9 months. This in turn creates pressure to pack as much functionality into the chips as possible (since it’s hard to predict everything that might be needed by the time products eventually reach the market) and requires large teams to work on design, simulation and test, all contributing to the hefty development cost.
So the current silicon paradigm is “big is beautiful” – big development teams, big chip designs, and big fabs to produce them.
Is big always beautiful?
There are several problems with this paradigm. Firstly, not every application requires super computing power; often simpler functionality is sufficient (or even preferable). Secondly, many use cases cannot support the high cost (both development and production) of silicon chips. And thirdly, the high capital costs and long lead times of the conventional semiconductor industry make it prone to significant supply/demand imbalances.
A good example of this can be seen in radio frequency identification (RFID). In the apparel market, RFID has been proven to decrease overstocking and improve both supply chain efficiency and top line sales. It is already deployed on more than 10 billion items every year, but is now being constrained by availability of silicon chips. Even without this constraint, silicon RFID chips are just too expensive to be used on mass market everyday items. So, for these fast moving consumer goods, we still have inefficient supply chains with poor traceability - if we could extend RFID use cases to these verticals, the benefits would be huge.
Beyond RFID, simple electronics could enable many other innovations to improve everyday life. Smart packaging could interact with consumers, providing useful information about the provenance and life cycle of their contents. Smart healthcare patches could improve wound healing and allow patient-led diagnosis or monitoring common health conditions, improving quality of life and reducing the burden on the medical system. Smart sensors could be embedded in buildings and other infrastructure to improve comfort and safety while also reducing environmental impact. The list is, almost, endless!
The potential market size of each of these applications is measured in the trillions – but only if the price is right. No matter how exciting the potential, they do not make sense at the cost of today’s silicon-based electronics.
Do something different
Paraphrasing Albert Einstein, if we want to address these new types of opportunities, we can’t just approach them in the same way as existing semiconductor devices. We need to stop and think about what these applications really need.
Firstly, cost is the most obvious and quantifiable metric. Alongside low development costs, we need extremely low production costs that are aligned with the nature of the items to which the electronics is being added. Today we think of a few dollars as being cheap for electronics, but in the context of packaging for fast moving consumer goods we need to achieve costs around 1 cent or less!
Secondly, within this cost structure, of course we need to achieve sufficient functionality for the required application – but without over-engineering the solution, as there may be no value from additional complexity.
And finally, form factor is also important. Most of these applications don’t really want a “box” of electronics, they want it to be seamlessly embedded within the product or packaging. So thinness and flexibility, along with appropriate durability in real-world environments, is required.
At PragmatIC Semiconductor we make flexible integrated circuits (FlexICs) that are thinner than a human hair, while avoiding the expensive machinery and materials required in conventional silicon chip manufacturing. They support functionality such as RFID, in addition to modest compute capability and a range of potential sensors.
The idea of using thin and flexible substrates is not a new one, but most other companies are focused on displays (which require a different set of technology choices) or on merely using conductive inks for flexible PCBs. There are also thinned silicon chips which have some level of flexibility, but these don’t address the cost challenge (in fact, thinning silicon increases the price, due to additional processing and incremental yield loss).
At PragmatIC we didn’t just develop the process for FlexICs, we also designed a scalable and modular production system – our FlexLogIC “fab-in-a-box” platform, which can deliver billions of FlexICs at a capital cost 100x lower than a silicon fab and with a production cycle time of a single day! As we expand production, we envisage hundreds of these systems distributed around the world, close to where the chips will be integrated into the end product, enabling much more resilient and localised supply chains.
Semiconductor innovation in Britain
Semiconductor manufacturing is important, but given how the silicon world has evolved, it is hard to see how a leading-edge silicon fab could now make economic sense in Britain (even if supported by significant government subsidy).
But, according to McKinsey & Co the most disruptive technology in the coming years is the Internet of Things, and the applications discussed above are a critical part of this technology revolution. By introducing a new way of creating ultra-low-cost semiconductor devices and enabling novel manufacturing models that avoid the necessary supply concentration of the silicon industry, we can take advantage of exciting new opportunities for economic value creation that also positively impact some of the biggest global problems we are facing today – including climate action, circular economy, sustainable agriculture and ubiquitous healthcare.
At PragmatIC we are very optimistic that, with the UK Innovation Strategy, this unique technology and production capability can expand and enable UK PLC to play an increasingly central role in the globally important semiconductor supply chain. We look forward to working with our partners and the new Business Innovation Forum to make it happen.
Author details: Scott White is CEO, PragmatIC Semiconductor