Saturday, July 24, 2010

TSR Market Report: Samsung Closing Gap with Ominivision

Japanese Tech-On published an article about the latest TSR report "2010 CCD/CMOS Area Image Sensor Market Analysis". Most of the article is quite hard to interpret (see Google translation), but the graph below apparently shows year over year volumes and market shares for major image sensor companies:

It looks like Samsung is a big winner this year with its market share approaching Omnivision's. SETi market share has stabilized this year, while Galaxycore's rise mostly comes at Omnivision's expense.

Via Japanese Image Sensor blog.


  1. My feeling is that the real rising sar is Galaxycore !

  2. Back in May there was a hyperlink to an I-Micronews article in which the Galaxycore founder freely admitted that his company succeeded because another company underwrote his company's research and development costs. In a follow-up post, Eric Fossum made some very reasonable comments about intellectual property piracy. Eric's comments drew a number of very angry anonymous replies. For the most part those replies didn't address the substance of Eric's comments, but relied on personal attacks to send a message that no one should dare criticize the success of this company.

    The impression I got from the May ad hominem activity was that there's a strong desire for the appearance of success to be acknowledged as success. I get the same impression from the first anonymous comment above.

    Eric Fossum is going to wake up this Sunday morning and read this blog thread, and almost certainly he's going to have an opinion some people won't like on whether two data points are sufficient to declare Galaxycore as the real rising star. He may share that opinion in rather blunt terms, and then we could be back to another escalating back-and-forth like the one in May.

    Before we get to that point (and hopefully we won't) I'll leave you with the following observation:

    If you can convince Eric Fossum that Galaxycore is a success, it will be a success the company will have well and truly earned.

  3. I've heard few theories of why Galaxycore is successful, other than Eric's IP exploit one. No matter if these theories are correct or not, Galaxycore is the real force in China that anybody should account for.

    Talking about the long term success, Galaxycore needs to compete on pixel technology with the more established players. So far Galaxycore is selling relatively large pixel low-end sensors. I think the real test would be if it's able to create the competitive performing 1.1um or smaller pixels. This would be the company's pass to the higher-end, higher margin market. It remains to be seen if Galaxycore can make this.

  4. Dear Image Sensors world,

    I think that could be the real marketing issue: do the general consumers care the pixel size ? Most of consumers will ask the price, image quality, they will never ask the pixel size in the camera chip at all.

    In China, the mainstream mobile phones use VGA level sensor, so the pixel size is not a issue taking into acount the low price of CMOS wafers. This is the real explanation of the low cost of Galaxycore sensor. Just imaging today that you use 0.18um or 0.13um CMOS to design a average quality VGA level sensor with 3.6um pixel size, what kind of R&D investment do you need except the mask set cost ?

    But when you go to 1um pixel with BSI, you have to invest a huge money inside. Is this justified by a VGA low cost sensor ? I don't think so. Make things simple and make things cheap, that is the key to success in China market.

  5. You are right, there is no need in small pixels in VGA world. However, margins in VGAs are 20% at best. To get better margins one needs to go for higher resolutions and smaller pixels. Higher resolutions are also needed to capture market outside China.

  6. If you can capture the majority of low end business in China, you are already the king !

  7. TSR does a pretty good job with this kind of analysis so my first comment is that 2007, 2008, and 2009 data is probably accurate. 2010 estimates could go differently from this although most 2010 business is probably booked. Nevertheless, if you cover up the 2010 estimates, the story looks a little different.

    To my eyeballs, it looks like Galaxycore is eating Omnivision's business, and Sony is eating Aptina's business but I am sure the situation is more complicated than that. For obvious reasons, I am glad to see the positive trend for Samsung.

    Galaxycore does not look like more of a rising star than Sony or Hynix to my eyes but it seems that we will know the answer to that more fully in a couple of years. I imagine the China market will become more sophisticated over time and I guess Galaxycore needs to put its money into R&D.

    The fact that we are talking about Galaxycore at all means it is being successful. We don't need to argue about that. But, does it earn the technical respect of the image sensor technologist community? Not yet, I think.

    But once upon a time, I think Matsushita, now Panasonic, was called the "copycat company" (relative to Sony). Yet, today it has grown into a well respected organization. Perhaps someday we will say the same about Galaxycore.

  8. I don't think that Galaxycore will move to high end products that fast. In general, chinese companies put much less money in R&D than japanese companies. Someone said that the goal of chinese company is "make money" and the japanese company is to make product. This means that if the low end products can bring reasonable profit, they will not move to the high end product. This phenomena can be observed in all fields.

  9. @ Sony is eating Aptina's business

    This brings another interesting question - where is Sony on the graph? There are three lines named in Japanese only and there are 4 companies in Japan: Toshiba, Sharp, Panasonic and Sony. Could somebody help to translate this?

  10. Hi,

    The blue curve below Galaxycore corresponds to Sony. Light blue below Aptina is Toshiba. Sharp is light green below Magnachip. I don't know for the last, below Pixel plus. If it's Matsushita, the translated Japanese form is different.

  11. There's an English translation now:


All comments are moderated to avoid spam and personal attacks.