June 18, 2019 | Jen Bernier
For those people who enjoy a glass of Cabernet, Sauvignon Blanc, or Pilsner, you know that the temperature of that beverage is key to its...
We’ve previously looked at the technical benefits of using custom integrated circuits when you’re designing, manufacturing and supporting devices destined for the Industrial Internet of Things (IIoT).
Something we only briefly touched on in that piece was the financial side, noting that custom silicon can be cheaper than the traditional approach of combining tens or even hundreds of discrete components on a circuit board.
Given that cost has long been seen as a barrier for equipment makers when it comes to custom silicon, we wanted to look at this in a bit more detail and show you how to work out whether bespoke is the right approach for your specific IIoT product.
Technology and manufacturing techniques have moved on a lot in recent years, so even if you’ve previously analyzed the costs of bespoke silicon vs integration of commercial off-the-shelf components, it’s worth revisiting: you might be surprised at what you could save. We’ve seen customers achieve savings of up to 80%, or deliver payback within 12 months.
What sometimes gets overlooked, when comparing the cost of custom silicon with the traditional approach of combining discrete components on a board, is that you need to consider more than just the bill of materials. With the traditional technique, you’ve also got the costs of sourcing each component, storing sufficient inventory, assembling the pieces and testing them. And that’s before one of your suppliers discontinues or changes its product, at which point you’re faced with some of this expense again. By investing up-front in custom silicon, you can minimise or even eliminate many of these traditional costs.
Next, it’s important to think about the total cost over the period for which you expect to manufacture the product. In the case of the IIoT, this will typically be longer than with most consumer electronics devices, which tend to be produced for a couple of years at most.
To work out whether custom silicon is likely to result in savings compared to traditional methods, start by thinking about how many devices you’ll be making each year. Next, consider how long you expect to be producing devices that use this chip (or variants based on the same design).
The next factor is the number of data converters you’d have on your traditional board, including those in your microcontroller(s). Then think about the number of diodes, transistors, regulators and other analog components.
Next, do you need a processor, and how powerful must it be? Are you talking a basic 8-bit microcontroller, or something meatier, such as a 64-bit multi-core ARM device? What about storage and RAM? Again, take into account the memory on your board and in the microcontroller.
Finally, do you require wireless connectivity? For an initial cost projection, it’s enough if you can simply answer ‘yes’ or ‘no’ to this question; the specific technology makes relatively little difference.
Armed with this information, head to our free calculator and plug in the numbers to see your expected break-even volume and total savings, along with a graph showing the comparative costs over time of custom silicon compared to traditional techniques.
You might be surprised by the results.