When the chips are down
SCOTT JONES, Alix Partners, San Francisco, CA
There's a need for greater R&D efficiency in the semiconductorindustry.
It's well known in the semiconductor industry that R&D is the lifeblood of the business. In 2012, R&D spending in the $300 billion chip industry neared $50 billion annually???roughly 16% of revenues. These high levels of investment are necessary to keep pace with the demands of Moore's Law. With long development cycles, these companies must place huge bets on technology trends years before the products they enable enter the marketplace. In this industry more than any other, the ability to align ones' R&D capabilities with market opportunities and deliver these programs to the market on time is crucial to the financial success of semi companies. In the industry study AlixPartners recently released, we show how powerful the lever of increasing R&D efficiency can be to increasing profitability.
Currently, the ability to fund the necessary R&D to fuel industry growth is under pressure from a combination of slower revenue and growth and rapidly rising costs for leading edge development. Our study shows revenue growth is slowing across the industry, pressuring R&D teams to maintain the same level of innovation with fewer resources. At the same time, R&D costs are increasingly dramatically???by as much as 60% for fabless chip designers for development on the leading edge. As a result, companies face a significant risk that the combined pressures of reduced R&D budgets and higher R&D costs will drive an innovation death spiral, in which an inability to fund R&D programs causes companies to continuously miss product cycles, leading to further revenue declines and market share erosion. In this environment, getting more out of your investment through improved R&D efficiency is critical.
Slow revenue growth threatens R&D investment
The rate of revenue growth in the semiconductor industry is near a 20 year low (FIGURE 1). Even as the industry rebounds from the trough in 2009, the recovery has not yielded sustained annual growth near the historical levels of 8 to 10%. The five-year rolling average has been in a range of 1 to 6% for the past four years and is expected to improve only slightly in 2013. While we expect a slight recovery in the market in the second half of 2013, growth is unlikely to reach levels that exceed 5%, well below historical averages. The industry average revenue growth rate has slowed to 3% since 2006, after averaging 12% growth from 1991 to 2006. As companies attempt to maintain R&D spending as a percentage of revenue, the slower top-line growth pressures R&D teams to innovate with lower R&D budget growth.
|FIGURE 1. The rate of revenue growth in the semiconductor industry is near a 20-year low.|
One of the dynamics that is influencing the market growth is the convergence of the PC market with mobile handsets through the growth of smart phones and tablets. Since the introduction of the iPad in 2010, the tablet market has grown rapidly and is cannibalizing demand for the traditional notebook PC market; while some of the demand for tablets has been incremental to the overall market, most tablet sales have come at the expense of the notebook PC market. The growth of the combined notebook and tablet market is marked by both a flattening of the growth trend for notebooks and an explosion in tablet growth since 2010. In 2011, tablet sales grew by more than 250% year over year while notebook sales grew by only 2%. In 2012, tablet sales are estimated to have grown at a 67% rate; notebook sales growth remained at a 2% rate.
Leading-edge R&D spending rising rapidly
Revenue growth may be slowing, but the same cannot be said for the need for and cost of R&D. The cost of developing leading edge products and chip manufacturing processes is increasing at higher rates each generation. A recent presentation at the Common Platform Technology Forum stated that integrated device manufacturers (IDMs) and foundries are experiencing 35% cost increases from one process technology to the next while fabless semiconductor companies are seeing even greater cost increases???of roughly 60%???for designs and tapeouts on next generation process technologies (FIGURE 2).
|FIGURE 2. IDMs and foundries are experiencing 35% cost increases from one process technology to the next while fabless semiconductor companies are seeing even greater cost increases???of roughly 60%???for designs and tapeouts on next generation process technologies.|
Leading-edge fab costs have increased at an average rate of roughly 30% per generation over the past decade and we expect at least the same sort of increase at 14nm, if not greater due to the complexities of moving to new transistor technologies necessary at geometries beyond 20nm.
The R&D costs necessary to develop a next-generation process manufacturing technology are increasing at a similar rate; at nearly 30% per generation, a rate that has remained relatively steady for the past decade. This is mostly due to the high cost of the test wafers and development tools necessary to deliver a new process technology. More expensive test wafers and tools mean that foundries and IDMs cannot afford to invest in as many customer specific or variant technologies. This puts more pressure on foundries to deliver a leading-edge technology platform that can address a sufficient segment of the market to allow for a positive return on their leading edge investment.
Chip design for leading-edge products is seeing the most rapid rise in costs. The cost to design a chip for a leading-edge process has increased at nearly 60% per generation over the past decade. This is twice the level of the increase in costs for leading-edge fabs and new process technologies. As a result, fabless semiconductor companies can afford fewer chip designs on the leading-edge processes. This makes it even more critical that the leading edge development programs have the features that end market customers require and are targeted at the market segments with the highest value. The upstream impact on the foundries is that fewer designs on the leading edge lead to a higher concentration of volumes on a small number of products.
Importance of R&D efficiency and time to market
The major risk for companies during periods of slower revenue growth is that pressure on R&D budgets will prevent the product development necessary to spur future revenue growth, leading companies to miss major product cycles. This creates a potential innovation death spiral for companies, in which an inability or unwillingness to fund necessary R&D programs leads to further market share erosion and greater revenue declines.
How does this spiral downward begin? When a company has missed market signals and its products are not aligned with end market needs, it is likely to lose key market share or to become too heavily exposed to a shrinking market (for example, notebook PCs). This can lead to obvious negative impacts on revenue and earnings. A common reaction to this scenario is to constrain R&D spending as part of overall retrenchment, but if this is not done in a careful and thoughtful way, the company risks accelerating its losses through continued market share declines resulting from a lack of innovation through insufficient R&D investment.
One of the biggest impacts of R&D efficiency is the ability to get products to market on time. For leading-edge products where power and performance count most, time to market likely means the difference between success and failure. The challenge for companies is not only to deliver individual programs on time but also to balance a portfolio of R&D resources and deliver multiple programs on time to the market???and to do it generation over generation. Often, organizations take on too many programs to gain market share on the current generation and starve resources for the subsequent generation, leading to an inability consistently deliver to market needs. The result of this R&D over-commitment is often that the effort to capture incremental opportunities puts the delivery of programs serving the core business at risk.
Efficient R&D leads to greater profitability
In a recent AlixPartners' study of the 72 largest publicly traded semiconductor companies globally finds that companies that are able to grow their revenues at a rate greater than R&D spending increases achieve increased profitability. As shown in FIGURE 3, thirty-four out of 41 companies???or 82%???that grew R&D spending as a percent of revenue more quickly than they grew revenue saw a decline in operating margins. But among those companies that grew R&D spending at a lower rate than revenue growth, 24 companies out of 31???or 77%???saw an increase in operating margins.
|FIGURE 3. Companies that grew R&D spending at a lower rate than revenue growth saw an increase in operating margins.|
This does not mean that companies should reduce their R&D budgets to lower levels to reflect their revenue growth. It does, however, indicate that companies that achieve higher returns on their R&D investment, as demonstrated through higher revenue growth, are much more likely to demonstrate higher levels of profitability over time.
Companies with poor R&D efficiency often suffer from a lack of proper planning more than an inability to innovate. The areas of highest importance are: aligning R&D investment to the market needs, allocating the proper amount of resources, and setting sufficient financial requirements for the programs to meet. Aligning R&D capabilities with opportunities prioritized from the marketing organization can enable management to create a list of programs that the organization is capable of delivering. After financial requirements and analysis have been provided, management should be able to rank the potential programs based on the strategic priorities of the company and on expected ROI. Once the execution of the programs begins, the key is to ensure that the highest rated priority programs are fully resourced and brought to market on time with all of the key features necessary to make the offering successful.
Innovation is a key competitive differentiator for semiconductor companies. Getting to market on time with the right products can make all the difference between success and failure. As revenue growth remains low and the cost of R&D continues to rise, the amount of money companies can profitably spend on R&D is pressured. In this context, optimizing the efficiency of the R&D investment process is of paramount importance. Successful semiconductor companies must improve R&D ROI through proper planning processes and better allocation of resources. In this way, companies can achieve top-line growth while lowering R&D costs per program, paving the way for future successes.
SCOTT JONES is a Vice President at Alix Partners. Previously, he was an equity analyst with JPMorgan covering the semiconductor sector and spent seven years with Intel Corp. in finance and operations.
Solid State Technology | Volume 56 | Issue 3 | May 2013