Big in Japan

The recent launch of Catapult’s new SystemC support at EDSFair in Yokohama was also an opportunity to confirm a long lasting EDA trend: SystemC is big in Japan.

Since SystemC’s debuts a decade ago, Japanese companies have always been at the forefront of the user community and, consequently, constituting an important market for ESL products. But acknowledging the fact doesn’t do us much good if the reasons are not understood.

So why is it that we have to look east to find the largest group of SystemC users? A number of explanations have often been suggested. The willingness to leapfrog US rivals relying on the RTL abstraction was clearly an important factor. Moreover, this move was facilitated by the reputedly top-down management culture in Japan companies. And to close the loop, the very strong referenceability in the local market accelerated the movement: after a few companies showed the way, the rest of them followed in a desire to not be left behind.

Preparing Recommendations

But if this helps understanding Japan’s early start, it doesn’t tell us why other regions, despite growing SystemC adoption, haven’t closed the gap. I believe that an explanation for this lies in a simple yet often overlooked fact: the kinds of designs done in Japan are different from the ones done in the US, which are also different from the ones done in Europe. Japan is at the forefront of multimedia design; US are unrivaled in networking applications and processors while Europe leads in wireless communications.

Am I simplifying? Certainly. Am I oversimplifying? I don’t think so. ESL requires domain specific tools and languages. Wireless modems are not architected like multimedia codecs. In second of part of this post, I am taking a deeper look at these differences and how they impact methodologies.

Tokyo's skyline from Mentor Graphics' office in Shinagawa

Preparing Recommendations