Wednesday, November 18, 2015

Samsung Updates on Image Sensor Business, Names RB-W Sensors BRITECELL

Samsung Investors Forum held on Nov. 16, 2015 gives interesting details on its image sensor business from Kyushik Hong, VP of Marketing, Samsung System LSI:

"But for some other product like image sensor, we have a number one market share in China market, for example. So, we sell a lot of our image sensor in all the Chinese OEMs."

Q: [ph] Jason (58:35) from [indiscernible] (58:34). "Can you please comment on [indiscernible] (58:38) how your image sensor technology compares to one of your Japanese competitors, and for the image sensor, besides mobile, what areas are you focusing on and how do you see your business penetrating into the mobile side?"

A: "Okay. So, I think for image sensor technology innovation. It has to come from, I mean – in large probably has to come from the process technology. So, having you around fab and you around process – I mean, the pixel process technology this is very important. So, I think that's not some fables, the base, the image sensor company can do. But particular company that you mentioned, a Japanese company, they also have their own process technology to develop the pixel quality. So, in that sense, I think we are pretty much on par.

But the one we have in addition to that is we also have very advanced logic technology as well. And probably you know that all the advanced image sensor is now all this [ph] packed (59:56) sensor. So, meaning that [indiscernible] (59:58) part of the sensor is surely the pixel and you put all this kind of logics that can handle all the immediate processing, stuff like that, into the separate dedicated logic chip and you just stack them together and these are too part of the sensor can work in perfect harmony as well.

So, Samsung, I think, one of the advantage we do have is that we do have both pixel [ph] in-house (01:00:27) pixel technology, as well as advanced logic technology as well. So that we can stack them together by taking advantage of the synergy between this technology.

And I think the second question is about the application outside the mobile, right? So, that's definitely another very exciting subject. Today's subject is about mobile, so I'm pretty much focused on mobile application. But we do have – we do see a lot of opportunity in other applications as well. And automotive is one clear example and, again, combining some of the processing power that we can provide.

So, for example, like automotive, like – there could be like 16 images sensors that has to be working at the same time, and you can't – you have to handle all of this kind of – the video images at the same time for very sophisticated [indiscernible] (01:01:33) type of automotive application. So, definitely automotive is one area. And there could be a lot of even consumer electronics area. Like in this case, even this robot [indiscernible] (01:01:49) they have some imaging sensors inside. So, we're constantly looking at some other opportunity. And we believe all the technology that we built upon the mobile device will greatly help to – for us to expand our area outside the mobile.
"

6 comments:

  1. From the small amount of modelling I did on this topic for Samsung, published at 2011 IISW, it seemed to me that RG-W would work better than RB-W. Perhaps better subsequent modelling showed the opposite. Anyway, the eye is more sensitive to G error than B error, hence better to sample G than B...plus G helps the luminance signal more.

    ReplyDelete
    Replies
    1. I always wondered about this. RG-W needs to calculate W-(R+G) to get B. Subtracting high value signals to get the low B signal.
      If that subtracting is not a problem, why not use CMY filters where 2/3rd of light reaches the sensor? I always thought because the subtraction would cause more noise, not less?

      Delete
    2. Subtraction and reduced SNR is definitely an issue in all these schemes. I also looked at CMYW. It seemed to be better than RWBW but not as good as RWGW. There are a lot of trade offs and optimizations to consider. Also, IP to work around or license. I also heard that fine patterning of CYM filters was more challenging than RGB filters, possibly due to maturity of the materials and processes.

      Delete
  2. No one raised the question re. the wide spread rumor that Samsung exited system camera business?

    ReplyDelete
    Replies
    1. Samsung just slows down developments of its camera lineup... in order to throw all its strength on the design of these new RB-W sensors (> +1 Ev)... then launch a new camera lineup with these RB-W sensors inside...

      Delete
    2. RB-W has been around for a while, including Aptina and TrueSense. Hope to see it materialize
      http://image-sensors-world.blogspot.com/2014/03/cfa-architectures-review.html
      http://www.anandtech.com/show/7149/aptina-announces-ar1331cp-13-mp-cmos-with-clarity-plus-we-take-a-look

      Delete

All comments are moderated to avoid spam.