Magma: a mirror on EDA
Last week Dylan McGrath wrote an interesting article on the problems facing Magma Design Automation. You can find it at Analysis: What's wrong with Magma?. Both Gary Smith and Roy Jewell, Magma's President and Chief Operating Officer, point out that semiconductor companies are not willing to purchase EDA tools on the basis of expected requirements. In spite of the optimism expressed by a few CEOs of EDA companies, Wally Rhines comes to mind, the demand for EDA tools with high license prices, and thus delivering the greatest profit margin, is softening and will not increase in the short term. Thus companies like Magma who depend on selling licenses for RTL to GDSII tools, will find it hard to sustain much revenue growth.
Although it is true that demand for electronic products, by an ever increasing consumer population in Asia, will keep the growth of end user markets at a healthy rate, the vast majority of those products will be lower end products that do not require the use of the latest silicon manufacturing technologies. As a consequence semiconductor companies will continue to be conservative in their purchases, while the technical problems to be solved by EDA vendors will increase in difficulty.
In the same article, Rich Valera, an analyst at Needham & Co., observes that Magma customers using Blast Fusion and later Talus products, used to enjoy as much as a six months advantage in delivering their designs to production over their competitors. But he says, the advantage is fading. Gary Smith agrees that the gap, not surprisingly, as narrowed, by he believes that the same customers still have a two or three months advantage. Given that many electronic products today have a market life of nine months, it seems to me that three months is still a very desirable advantage.
Obviously the EDA market, that has always been very competitive and price sensitive, is becoming even more so, and the poor financial results reported by Cadence, Magma, and Mentor reflect it.
I agree with the need to expand in new markets, Magma has just entered the analog and mixed/signal markets with what seems to be well accepted products. But I think that all EDA companies are being too parochial in limiting themselves to EDA. Design Automation is a much larger market, covering a number of disciplines with problems similar to those faced by electronics engineers. What ever happened to thinking outside the box?