In this blog post I want to share some thoughts on the journey from product idea to revenue, to make it easier to understand the life cycle for a hardware component supplier in general and for Sivers IMA in particular. I will also touch upon how one of many 5G revenue streams can develop according to available research.
As we have shared in many quarterly calls and via press releases Sivers IMA is now entering the revenue phase for our unlicensed 5G product portfolio (see picture 1 below) and will as planned be in that phase for our licensed 5G products in early 2020. It is not always easy to understand how fast or slow a full product development and commercialization cycle is for a hardware company. It is also very different if you sell products to used in consumer electronics (e.g. mobile phones), wireless infrastructure or cars (automotive). Sivers IMA wireless business is today focused on wireless infrastructure hardware within 5G, which will be the main focus of this blog post.
Many years ago, when I was working for Sony Ericsson (SEMC), we were developing smartphone products, our design cycle was from approximately one up to three years from first idea to commercial market launch. Of course, some of the components, like the application processor had a longer development cycle. And to be included in the phone it had to be ready at least in some shape or form earlier to be selected for being designed into the phone (i.e. what is called a “design win” in this case for the application processor provider). The development cycle time is significantly affected if you are reusing a “product platform” or if it is a completely new product.
At SEMC new platforms was called a “Mother” and other new product based on the same platform was called a Daughter. Developing a Mother compared to a Daughter resulted in quite different development cycle times. The same type of platform concept is used in the automotive industry, which is known for quite long development cycles. Even if automotive companies nowadays have adopted the platform thinking and thereby reduced their cycle time over the last 20 years. However, a new automotive product is seldom less than three year from start to finish, often closer to four years or more.
Picture 1: Sivers IMAs “Mother” product. Which is entering the revenue phase for our unlicensed 5G products. The IEEE award winning unlicensed 5G chip – TRX BF/01 57-71 GHz RFIC, is mounted on Sivers IMA BFM 06010 RF antenna module, making it a complete RF unit/Module
To make it more specific, I will explain the overall development cycle at Sivers IMA. We make RF chips and antennas for the wireless infrastructure market and the four main steps in our product development cycle are listed below:
- Prototype development phase (8-14 months): Product specification, system partitioning, design, layout and fabrication of the prototype chip (or break-outs to test different blocks before you make a full chip). The time is of course dependent on the complexity of the chip and the foundry used for fabrication of chips etc.
- Test and verification and pre- qualification phase (3-5 months): Customer testing of prototypes to find issues and bugs and qualification of products.
- Final product development phase (6-10 months): Redesign and fixing issues and bugs, fabrication of the “final chip”.
- Re-test and verification phase and final qualification phase (3-4 months): Final verification and final product qualification.
Total cycle: Approximately 2-3 years for a “Mother” product including RFIC development
Our customers development and the commercial cycle
In early stages our product development cycle runs in parallel with our customers’ development and commercial cycle, which for a component company is connected with the product development cycle of the system vendor and its cycles.
The main steps in the cycle for our customers are similar to what I described above for Sivers IMA, even though they can often remove development time, since they base their products on “standard electronics” and often do not have to deal with the longer lead times related to silicon wafer manufacturing. In most cases, the development cycle time for our customers is a 9 to 18-month cycle including market approval and customer trials and maturity of the products.
The customers’ development cycle starts in parallel with our development, i.e. often during customer sampling of prototypes in step 2. During this phase Sivers IMA start getting design wins for a new product. As you have seen in a number of press release from us, to date we have nine design wins that has has been announced.
A design win means that a customer has signed a contract and has decided to develop a product based on our 5G technology. To reference one example, below to the right you can see a picture of a ready product by CCS, the 60 GHz Metnet Node, which is using the Sivers IMA BFM 06010 RF antenna module with the TRX BF/01 57-71 GHz RFIC. This product is now commercially launched and has started to be deploy, in for example the UK.
The second product, the CPE below to the left is a “Daughter” to the Metnet Node and will be commercially launched this quarter (Q2 2019). This shows the full cycle from unlicensed 5G chip product idea to our customers finalized product that has now been launched and are deployed in the market.
CCS CPE home unit CCS MetNet Node
The effect of the commercial cycle
Sivers IMA “BFM 06010” RF antenna module with the “TRX BF/01” 57-71 GHz RFIC is already used in customer products and Sivers IMAs “Daughter” product the licensed 5G chip – TRX BF/02 24-29.5 GHz RFIC is now in prototype phase and will be ready for product launch late 2019/early 2020.
Our customers, the system suppliers (the second step in the value chain) are often ranked based on size, e.g. Tier 1 or Tier 2 etc. For example, Huawei, Ericsson and Nokia are typical Tier 1 system vendors, while CCS for example can be ranked as a Tier 2 including their partner ADTRAN. The same is valid for the mobile operators that are customers to the system vendors, they are also ranked by size in Tier 1 to Tier 3. For example, two of the biggest Tier 1 mobile operators are Verizon and AT&T. The bigger they are they more units are expected to be sold and delivered.
Evaluating anticipated revenue in an emerging market like 5G requires the use of external reports and resources. To focus on one example, I like to share data for the FWA (Fixed Wireless Access) market. It is the first mmWave 5G vertical that is expected to grow into a larger market. Other verticals such as transportation (trains) and medical are expected to add revenue as well the coming years, but are not part of the below estimates.
For this blog post I use the SNS Research report “5G for FWA (Fixed Wireless Access): 2017 – 2030” (www.snstelecom.com/5gfwa). This report estimates approximately two million RF units to be shipped in 2019 and a total of 225 million units until 2028 (with 60 million units per year in the year of 2028). The 5G market is something that will grow over a very long time and there will be big investments, for example just recently is was announced in Japan that the 4 operators will spend USD 14 billion in 5G infrastructure from 2020 to 2025 and there is a 28 GHz band available which offers 400 MHz wide channel for mmWave 5G.
If we apply a unit price per RF of USD 15-70 (depending on how complex the RF units that are used), the Total Addressable Market (TAM) for RF units should have a value between USD 3,3 billion and USD 15,7 billion. for this time period. For simplification and to count on the lower side, let’s assume a TAM of USD 4 billion.
In total the Tier 1 system vendors will jointly typically reach a market share of 60-75 percent in a vertical like FWA, the rest of the market will be addressed by Tier 2 and 3 companies. Let’s assume that over time there will be approx: 50 vendors in total for this specific market segment. Then you can divide the total market in three parts, 60 percent to Tier 1, 30 percent to Tier 2 and 10 percent to Tier 3. This simplified estimate gives this table:
Estimate of RF vendors revenue per system vendor for mmWave 5G RF units for FWA (2020-2028):
- Tier 1: TAM USD 2.4 billion with 5 Tier 1 this is ~ USD 500 million per vendor
- Tier 2: TAM USD 1.2 billion with 15 Tier 2 ~ USD 80 million per vendor
- Tier 3: TAM USD 0.4 billion with 30 Tier 3 ~ USD 30 million per vendor
This gives a possible view of the overall expectations and a possible value for each design win from 2020-2028 for a “hypothetical” standardized Tier 1, 2 or 3 customer.
In summary, new products within the wireless infrastructure sector will start generating volume product revenue for a hardware company like Sivers IMA, after three to four years. Sivers IMA’s wireless unlicensed 5G product portfolio has been through 3.5 years of development, and are entering the volume phase for FWA. This is just one example in other verticals like medical or transportation design wins can have quicker or slower cycle . For cars within the automotive industry you need to add maybe even two to three years on top of this, to get to the same phase as with FWA for infrastructure.
Hence the number of design wins with the right customers are one of the most decisive factor for the total outcome, that is directly connected to the success of the customers’ products in the market. From the above product development cycle times, as well as this simplified commercial market revenue model, it is possible to make assumptions how and when a component supplier will achieve commercialization and volume revenue. These numbers are of course somewhat simplified and estimates based on the 5G scenario by SNS Research that you have to evaluate for yourself and believe in or not. Even though this is an example and in someways is hypothetical, I hope it offers in insight into the total time to market for a hardware company from product idea to market commercialization and finally revenue and I hope it will help you as an investor or analyst to understand this journey from product idea to revenue.