Design of electronic appliances gives you a wide range of design options. Each option has its pros and cons. You will often realise that the design is not limited by electronic parameters. It is more likely limited by power, type of sensors, communication bandwidth, space, and form factors you have at your disposal and – last but not least – your product’s market sales price and thereby your product manufacturing cost.
The optimal solution is like a Gordian knot. It is a compromise between many factors mostly driven by your access to skills and resources. Knowing the exact volume and product life time would be an ideal scenario. However, more often the basis of your decisions is mere predictions.
Many projects start at demonstrator level by use of standard components (off-the-shelf Integrated Circuits – IC). When the demonstrator or the first series is accepted, the cost must be cut down in order to meet market expectations. Other projects start with a product idea that requires high integration due to the form factor (e.g. portability) or available supply of power. These are the typical scenarios where it is relevant to evaluate whether a custom made integrated circuit (also called Application Specific Integrated Circuit or ASIC) is the right choice.
What is an ASIC?
An ASIC is an integrated circuit (IC) made for one specific purpose – a custom made chip so to speak. This is quite the opposite to a standard electronic component where you buy off-the-shelf ICs and only use part of the functions on the chip.
Most integrated circuits consist of a “die”, a lead frame or substrate, and a package to protect the circuit. A die is semiconductor material with a functional electronic circuit on board – see Picture 1.
An ASIC is designed and manufactured as any other IC. However, it is 100% focused on the specific application it is made for. This means that the ASIC is optimised for performance and thus eliminates the redundant functions, area and cost that is unavoidable when using a standard component.
A standard IC is general-purpose electronics i.e. designed for a broad range of use. The IC is designed in exactly the same way as an ASIC, by use of the same technologies, libraries and design tools.
The standard component is excellent for fast prototyping and to obtain a short time to market. The main drawback is that when you ramp up in volume, the only advantage you get is a small price reduction. If you need several standard components to do the job, you will have a lot of devices to keep track on.
Another drawback is that you do not know for how long this standard component is on the market. Once a standard component becomes obsolete, you need to redesign your product. This is often critical for a company as it steels precious resources from your other important design tasks and creates a vulnerable gap in the product supply.
An ASIC gives you full control of the supply chain and you avoid nasty surprises raised by obsolete components.
|Time to market||++++||÷÷||Standard components are ideal for prototyping and small series.|
|Non-Recurring Expenses (NRE)||+||÷÷÷||Introduction cost is low for standard component applications.|
|Unit cost||÷÷÷||+++||With the optimal set-up for a given volume the unit cost for ASIC is very low especially due to no redundant functions and overhead|
|Scalability||÷÷÷||+++||With an ASIC it is easy to increase the volume without extra effort.|
|Power consumption||÷÷||++++||The design for an optimal power consumption is the real strength of ASIC technology with no redundant functions and dedicated technology for the application.|
|Form factor||÷÷÷||++++||ASIC is ideal for portable and low power applications due to its small form factor and power consumption.|
|Control of supply chain||÷÷÷||+++||Transparency in the supply chain: Access to wafer fab, assembly and test house is provided with the ASIC supply chain. For a standard component the distributor is the only source.|
Table 1: Comparison between a standard component versus an ASIC
Best performance and lowest unit cost & power consumption
The choice is simple when full custom design, large volume and small form factor are required. Large volume and smallest form factor equal lowest price for ASIC.
Optimised design enables best performance, the lowest power consumption and the smallest form factor. However, there are other situations where an ASIC is relevant. We will in the following address various relevant parameters.
With integration it is obvious that the dimension of the electronics shrinks. What is often not so obvious is that the packaging of the electronics can be optimised for an even better form such as low component height which is standard in mobile electronics such as phones and tablets.
Ensure your production
A product which consist of several standard components and has a long product lifetime will inevitably need a redesign when one or more standard components become obsolete.
A reduced Bill of Material (BoM) will prevent many challenges of redesign. Thus could an ASIC be the solution as an ASIC is an integration of many components into one single chip.
The challenge can be to control the entire ASIC supply chain. This is, however, one of DELTA’s core competences which we have handled for many years for many customers world-wide.
Protect your know-how
Integration of the electronic design into an ASIC first of all provides you with the smallest form factor. In addition, an ASIC also secure copy protection significantly because reverse engineering will be very difficult.
Cost parameters for integrated circuits
Most electronic components consist of a die, a lead frame or substrate, and packaging. More advanced components can have multiple dies and/or components in the same package.
Test and logistics (i.e. handling, shipment, quality assurance and control) are required to provide functioning components. For a typical IC the cost structure looks like this:
|Die||Technology, die size|
|Packaging||Type of package|
|Testing||Test time, yield|
|Handling||Logistics, shipment, QA/QC|
Table 2: Cost factors for ICs
The above cost structure is the same for a standard component and an ASIC. However, the cost of marketing, distribution and storage must be added to the standard components. This is the manufacturing cost structure, but what about the sales price?
Price structures for a standard component versus ASIC
The development cost (also called the Non-Recurring Engineering (NRE) investment) for an ASIC is high. The lead time is significant compared to buying an off-the-shelf standard component from a distributor.
The price of a standard component is low because many users split the NRE and together they reach a significant volume. The drawback is that distribution of components is a cost factor. In general, the gross margin for standard devices is 60%.
The price of an ASIC is defined by the specific volume that you as a customer forecasts and order. A typical gross margin is 20-40%.
When to choose an ASIC over a standard component
The obvious reply is “When the volume is high enough” (see Picture 2). In Picture 2 a standard component and an ASIC are compared in a graph where price versus volume is the only parameter.
However, there are many other parameters to consider such as: Do you need a small solution? Do you need low power? Do you need a long product life time? Many customers do ASIC implementation even though the volume does not meet the crossing point shown in Picture 2.
To sum it all up…
If an ASIC is the right solution depends on your specific project. However, you need to do an economic Return On Investment analysis and also to consider all the parameters below in order to choose what are most essential to your product:
Time to market
High versus low NRE
Big or small form factor
Degree of control of the supply chain
Protection of know-how
Do you have any questions? Do you need help to make a feasibility study to help you decide on your current project? Please contact us at email@example.com and we shall be happy to discuss it further with you.