Discussion: The IDM ROHM Aims to Be

Higashida

Most of what are considered major semiconductor manufacturers are now generally breaking away from IDM. To win out over those companies, I believe ROHM has to create an environment within the company where IDM can be a strength. On the other hand, we also need to strengthen IDM by utilizing external organizations like foundries and OSAT. ROHM sees making IDM a strength while utilizing external resources as a par tial breakaway from IDM.

Specifically, we will not make investments based on fluctuat-ing demand that have production costs that then become fixed costs, but we will make them variable costs by outsourcing to foundries and OSATs. This will also lead to maintaining profit. For example, we will make a management decision to outsource products that are not technologically unique and will not lead to future technological innovation, making them into variable costs.

However, even if we are outsourcing now, ROHM has technol-ogy in-house to create what we want to create whenever we like. The strength of IDM is that it can maintain in-house produc-tion efficiency according to demand and maintain costs. On the other hand, if we complete all processes in-house, we will have no experience with other companies' technology and will lack knowledge of the outside world. Utilizing outside resources is also important to acquire and study a variety of information.

Fujitani

On top of guaranteeing high quality, IDM can refine original production technology. That is also connected to ROHM's "quality first" corporate philosophy. Up to now, LSI has secured competitiveness with a sense of speed unique to IDM, in which the development cycle moves quickly in a way that closely aligns with customers. As for discrete semiconductor devices, we have acquired and maintained a significant market share thanks to the development of lines that can produce large quantities of good-quality products in-house. The IDM system can be further enhanced, but depreciation expenses from invest-ments made when demand was active become a large burden when demand slows down. So I agree that it is difficult to respond to the risk of fluctuating orders with 100% IDM.

In addition, development in IDM tends to veer toward products that are limited by the capacity and types of production lines within our own company, with the resulting risk that it is difficult to create new products. We have to temporarily outsource the manufacturing of products that can be stably produced, allocate our in-house resources to more important places where we need to hone our strengths, and pursue new information from outside the company too. A partial breakaway from IDM is also meaning-ful from the perspective of making sure that IDM does not con-strain or hinder the evolution of the development field.

Higashida

We should determine which parts of IDM to break away from depending on the circumstances. In the field of semiconductor miniaturization, especially, the production system of foundries is well-established in terms of scale. For example, in the future, we could change to a method in which the production volume of miniature semiconductors of 90nm or smaller would first be expanded through skillful use of foundry technology, then be strengthened with the in-house IDM system. Conversely, in fields where there are still no large foundries, like SiC semiconductors, there is the advantage of gaining sales as a pioneer by launching products with IDM, starting from R&D, to lead to technological innovation. Moreover, too much reliance on the outside will degrade the technology that is our strength and make it harder to distinguish our superiority over other companies’ products.

Fujitani

ROHM has competed using our own design and integral technologies. To return to such manufacturing, it is necessary for the future growth of the LSI business to secure the good parts of IDM while partially breaking away to delegate work externally for opportunities to gain new technology

I also think there is still room to seek cost reductions within ROHM for discrete semiconductor devices. ROHM’s in-house engineers have built an innovative production system that did not previously exist, so the launch of products on the market has been very fast, and we have also been ahead of our competition in reducing costs. However, there have been no dramatic improvements in productivity in recent years, and our overseas competitors are also catching up. I think that creating added value beyond expectations while curbing costs by adopting new materials and studying assembly techniques is something we should strive for in the discrete semiconductor device field in the future to compete with those overseas manufacturers.

Higashida

Since the emphasis tends to be on costs for general-purpose products, it is definitely difficult to compete head on with Chinese manufacturers. However, I think that we have a sufficient chance to come out on top if we differentiate ourselves in the aspect of added value, such as functions and abilities, by leveraging the IDM system. The international market is vast, and there are fields into which Chinese manufacturers still have not entered.

At the same time, we are always paying attention to the cost structure of Chinese manufacturers, and we are continuously comparing them to ROHM’s costs. The Chinese government’s assistance measures for the semiconductor field may not continue in the future, so I think we have to be sure to grasp our chances to win while ascertaining changes in the situation.

Fujitani

Regarding costs, it is also important to optimize efficiency in the manufacturing process. We verified a trial introduction of quantum technology solutions into the EDS process in our collaboration with Quanmatic in 2023. IDM requires fairly complex calculations to digitalize operations while anticipating every pattern, like facility installations and process variations, but the time for such calculations can be greatly shortened by quantum computers. It is groundbreaking that installations that used to rely on the skills of very experienced employees—one might even call them artisans—can now be reproduced with software. This technology not only reduces costs but is also expected to demonstrate effectiveness for saving labor and changing to unmanned processes in a country with a declining labor force.

In addition, our in-house engineers as human resources for manufacturing are also a significant strength of ROHM. From a human resource perspective, ROHM has what it takes to create world-first technology. In the future, the partial breakaway from IDM will lead to adopting outside technology, and the hope is that our in-house engineers’ development and proposal abilities will be even more refined, and we will once again be able to create world-first manufacturing proposals.

Higashida

ROHM was originally a company where IDM was launched vertically by business unit. However, if businesses are too vertically integrated, we cannot exhibit our strengths as an IDM. I believe that a matrix structure is ideal for the organization. That is why, from my point of view as the person in charge of the Wafer Process (WP) Production Headquarters, I have worked cross-divisionally to share technology to various divisions.Overlooking everything in this way shows that one division’s technology can be used in other divisions too.

Fujitani

I have also worked to optimize the overall production structure across business units while in charge of the Assembly Process (AP) Production Headquarters. I think this has produced certain results, such as eliminating the waste of overlap between businesses in the production system and leading to standardization, but the business system has now returned to vertical division to start rebuilding a strong business foundation. I think this is an important management decision from the viewpoint of optimal business.

Higashida

I feel the same. What ROHM needs now is to create a strong IDM. Vertically integrated businesses will actually show their strength remarkably if market conditions decline. That is why I think it is a very good thing that the company made a swift decision in response to changes in the current market conditions. It is important in management to maintain the matrix structure and flexibly determine whether to place importance on vertical or horizontal integration according to the environment at the time.

In fact, the strategic divisions for WP and AP remain as cross-divisional organizations. I would like to create an IDM that is unique to ROHM by leveraging the sense of speed that comes with vertical integration, supplementing where necessary with cross-divisional organization for overall optimization.

Fujitani

I would like to aim to become an attractive company like nowhere else, not a company one can find anywhere. Although fewer companies are using IDM production, I think it is still in the field of IDM that ROHM can discover a way out of difficulties. The partial breakaway from IDM, making efforts in that field without adhering to it too much, and doing some outsourcing, is also the embodiment of our policy of creating a stronger IDM.