Please note that JavaScript and style sheet are used in this website,
Due to unadaptability of the style sheet with the browser used in your computer, pages may not look as original.
Even in such a case, however, the contents can be used safely.
Volume 6 (Aug 25, 2003)
Facing the Challenge of CB-90 Development (2/2)
Part 2: Design Development
An unexpected pitfall
Senior Design Engineer
Tomoaki Isozaki
Three library types were prepared for CB-90 to meet user needs: the CB-90H, suitable for use with broadband communications equipment, servers and other GHz ASICs; the CB-90M for digital consumer product use; and the CB-90L for low-power-consuming mobile information terminals, and so forth. These libraries include various functions, such as several hundred to several thousand primitive cells (logic gates such as inverters and NAND gates that can be handled with CAD tools), I/O buffers and more. Clear-cut design rules were indispensable to the development of these libraries. Thanks to the painstaking work carried out by the working group for creation of the design rules and the working group for development of the library technology, development seemed to be proceeding relatively smoothly.
Nonetheless, there were of course problems. Putting together the logic meant lining up the primitive cells exactly according to the design, and this was done using CAD tools. However, with the delicate 90 nm process, a new problem arose that had never before been encountered. For some reason, the slits that occur when cells are lined up could not be kept within the established standard range (Figure 3). Although Tomoaki Isozaki, who was in charge of library development, at first thought that the problem could be solved by changing the way the cells were made, it was feared that doing so would have an adverse effect on performance.
Isozaki, who was very troubled over this, came to the conclusion that the problem could not be solved by merely focusing on the primitive cells. The person in charge of process development and the person in charge of tool development were able to solve this problem by sharing the task at hand and combining the countermeasures they came up with on their own, including reconsideration of the slit design rule, improvement of the tools used to line up the cells, improvement of cell creation without affecting performance, and so forth. After all was said and done, it took three months to find a solution to this problem. According to Isozaki, "If we had not been working in an environment in which we could freely consult with each other, I am certain that it would have taken even longer to find a solution and it would have been more costly as well."
The completion of CB-90: The ultimate goal was yet to be reached
On November 11, 2002, after its spin off from NEC Corporation, NEC Electronics announced its first product the 90 nm process for system LSI design, CB-90. This was the moment when the work that these engineers had devoted all of their energy to became a reality and the fruit of their efforts was unveiled to the world.
Once it went on the market, CB-90 received high acclaim. It even won the fiscal year 2002 "Nikkan Industrial Newspaper Top Ten New Products Award" and was the only product to win the award from the semiconductor sector. However, this was not the ultimate goal. CB-90 was subsequently taken over by a new team and great efforts were made to prepare it for mass production. Needless to say, the engineers didn't even have a chance to take a break before having to start their next challenge. With total capabilities from development and design to production, NEC Electronics puts to use all possible merits and takes on the challenge of next-generation development. The immense pride of its engineers is involved in the creation of all cutting-edge technology by NEC Electronics.
The 90 nm process is expected to become widespread through the utilization of STARC (Note 1), which was jointly established by major domestic semiconductor makers, and the ASPLA (Note 2) experimental production line, established with the objective of standardizing the 90 nm platform.
Note 1: Semiconductor Technology Academic Research Center (STARC)
The Semiconductor Technology Academic Research Center (STARC) was established by Fujitsu Limited, Matsushita Electric Industrial Co., Ltd., NEC Electronics Corporation, Oki Electric Industry Co., Ltd., Rohm Co., Ltd., Renesas Technology Corporation, Sanyo Electric Co., Ltd., Sharp Corporation, Sony Corporation and Toshiba Corporation, all of which are shareholders. Through collaborative research on the area of semiconductor technology, STARC aims to strengthen technological bases and preserve competitiveness.
Note 2: Advanced SoC Platform Corporation (ASPLA)
Advanced SoC Platform Corporation (ASPLA) was established by Fujitsu Limited, Matsushita Electric Industrial Co., Ltd., NEC Electronics Corporation, Oki Electric Industry Co., Ltd., Rohm Co., Ltd., Renesas Technology Corporation, Sanyo Electric Co., Ltd., Sharp Corporation, Sony Corporation and Toshiba Corporation, all of which are shareholders.
