You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
It has been reported in several publications yesterday that Samsung might spin out their foundry business due to the loss of Apple business. I wonder if the IP disputes between Apple and Samsung had something to do with this? Could this dispute have damaged Samsung's attempts to secure other business for their foundry? Are there any opinions of what one of the big three spinning out their foundry business will have on the industry? This was reported in Barron's, among other publications. Barron's is one of my main sources and I consider them highly reliable. Except for TSM, Taiwanese suppliers to Apple are reporting orders slowing. Any opinions or observations are solicited and welcome.
Samsung Electronics is considering a reorganization of the System LSI division in order to systematically grow the system semiconductor business. The company is planning to separate the design and manufacturing sectors in the business unit and divide or spin off it to fabless and foundry business divisions.
The consideration came after the company lost Apple, the largest customer in the sector of application processors (AP), which refers to the brain of smartphones, to Taiwan’s TSMC and some raised awareness towards the need to separate the foundry business division.
Samsung Electronics’ System LSI business division is largely divided into four segments; system on chip (SoC) team which develops mobile APs, LSI development team which designs display driver chips and camera sensors, foundry business team and support team. According to many officials in the industry, Samsung Electronics is now considering forming the fabless division by uniting the SoC and LSI development teams and separating from the foundry business.
The System LSI reorganization plan shows Samsung’s will to grow the business more systematically by dividing its system semiconductor design and manufacturing capabilities.
As Samsung Electronics succeeded in mass producing system semiconductor products 10-nanometer process technology for the first time in the world, the company signed a large foundry contract with Qualcomm.
I do agree that IDM foundries who compete with customers have a greater challenge but I'm not sure "spinning" a division out will solve that stigma. Remember, when the foundry business first started it was based on IDMs renting out excess fab capacity. Morris Chang was at TI and saw that as an opportunity for a pure-play foundry business model and the rest as they say is history. All things being equal, fabless customers will always prefer the pure-play over IDM foundry business model, my opinion.
Rather than spinning the foundry business out, if I were Samsung I would invest more in the ecosystem and do something disruptive. In fact, the first thing I would do is buy a fabless ASIC company like eSilicon or Open Silicon. Samsung buying Cadence would also be a hugely disruptive move. Free tools for all Samsung foundry customers!
The idea of a foundry buying a big EDA company has been around for a while (Mike Gianfagna often talks about this) and certainly has some appeal, but I would have thought there would be concern that once you chose to work with such a company you would be locked in - they're not going to let you use their tools to fab with someone else after all. This works better if the foundry buys just the back-end tools but that would mean buying a Cadence for example and then spinning out the front-end piece. Would that make sense to shareholders?
It has been reported in several publications yesterday that Samsung might spin out their foundry business due to the loss of Apple business. I wonder if the IP disputes between Apple and Samsung had something to do with this? Could this dispute have damaged Samsung's attempts to secure other business for their foundry? Are there any opinions of what one of the big three spinning out their foundry business will have on the industry? This was reported in Barron's, among other publications. Barron's is one of my main sources and I consider them highly reliable. Except for TSM, Taiwanese suppliers to Apple are reporting orders slowing. Any opinions or observations are solicited and welcome.
Samsung's business model is full of contradiction and conflict of interest. The relationship among their smartphone, LCD display, battery, memory, Exynos processor, SoC design services, and the contract foundry business is a perfect example.
They need to behave as happy brothers and sisters live under the same roof even though each of them knows very well that there are better suppliers or buyers outside of Samsung family. Those outside suppliers or buyers might offer better price/profit, quality/quantity, technology, performance, and availability. Yet when the global market is good, each of these Samsung units can't maximize their own profit because they need to take care the in-house customers or suppliers first. When the global market is bad and has too much supply, these Samsung units again can't cut down their cost fast enough by switching to outside cheaper suppliers.
And we haven't talked about all those conflict of interest Samsung's customers are facing every day.
All these contradictions and conflict of interest can't be solved unless each of these Samsung divisions can be a truly independent company.
I tend to agree, but good luck with seeing that happen. Samsun represents (last time I heards) ~1/3 of the South Korean GDP. A good object lesson in what happens when a company really gets too big to fail. It's not longer just a company, it's an integral part of the economy.
I tend to agree, but good luck with seeing that happen. Samsun represents (last time I heards) ~1/3 of the South Korean GDP. A good object lesson in what happens when a company really gets too big to fail. It's not longer just a company, it's an integral part of the economy.
The change will come except it may not come under Samsung's terms and under an ideal timetable. They can change themselves now when they still have certain ability to mange it. Otherwise the global market will force them to change but it will be much messy and ugly.
Dan, I agree completely. I don't ever see Samsung having a truly independent foundry business. Trust should always be earned and not given in a situation like this.
I wonder what they mean by "independant foundary". Would they spin off only the logic foundry, or would the DRAM and NAND business be included (because technically that's foundry too)
The idea of a foundry buying a big EDA company has been around for a while (Mike Gianfagna often talks about this) and certainly has some appeal, but I would have thought there would be concern that once you chose to work with such a company you would be locked in - they're not going to let you use their tools to fab with someone else after all. This works better if the foundry buys just the back-end tools but that would mean buying a Cadence for example and then spinning out the front-end piece. Would that make sense to shareholders?
Well, from what I know Calibre from Mentor is the signoff tool for many companies for the verification, Star RC for the extraction ... Cadence Virtuoso for the analog design. I think it's difficult for a foundry to fully qualify multiple DRC decks for sign-off on a given technology - same with simulators and RC extraction. My point is I don't think it's possible to do a full back-end design with the tools from one CAD vendor only today...
Dan, I agree completely. I don't ever see Samsung having a truly independent foundry business. Trust should always be earned and not given in a situation like this.
It makes sense to me they would spin off their foundry business -- as an independent company. (not affiliated with Samsung) Samsung would also lose some foundry business with key large companies because they wouldn't want to be dependent on their key competitors to provide important ICs. I think it is very clear.
It has been reported in several publications yesterday that Samsung might spin out their foundry business due to the loss of Apple business. I wonder if the IP disputes between Apple and Samsung had something to do with this? Could this dispute have damaged Samsung's attempts to secure other business for their foundry? Are there any opinions of what one of the big three spinning out their foundry business will have on the industry? This was reported in Barron's, among other publications. Barron's is one of my main sources and I consider them highly reliable. Except for TSM, Taiwanese suppliers to Apple are reporting orders slowing. Any opinions or observations are solicited and welcome.
My 2 cents: If this had been titled "Samsung to pursue equity offerings" that would sound likely to me. This is not likely to be specifically linked to Foundry though.
One day if average Korean people start thinking Samsung is just too big and it's threatening the justice system and the democratic society they are pursuing, then Samsung will be broken up.
