Friday, July 03, 2015

Press Covering Invisage QFAB3 Inauguration

Invisage publishes an impressively big list of newspaper articles covering the opening of its QFAB3 in Taiwan last month. The article that has attracted my attention is a Chinese site 52RD interview with Jess Lee, Invisage CEO. Few quotes, in Bing translation:

"Compared to CMOS, quantum films have a very big advantage, is very thin. Jess said, quantum well thin film sensor modules can do 4mm, and minimum can only do 4.8mm CMOS sensor module, though only the 0.8mm gap, but used on cell phone cameras and wearable devices, will be a big change, maybe your next-generation iPhone camera will not be highlighted."

"The Taiwan new factories are small, but the plant can support 10% market. Production processes are TSMC's wafers first, then we will spread the quantum film, then to partners to test, then made modules. This process and the traditional CMOS is the same, just adding the cover of quantum film, a step, so quantum thin film sensor camera performance can be increased 5 times."

"At present, we only produce a small number of wafer and will ramp up in the second half of this year. ...our first samples had been provided to a number of important customers, they have done tests and gave very positive feedback, later this year we will have more samples available to more customers. Over time, our formal product launches will be held in Shanghai, and after product formal release, we will provide samples for Mainland, 7 August should be, we can look forward to you."

"InVisage will focus on cell phone cameras, this is a very large market and opportunities. Smartphone carrying quantum thin-film technologies are expected to be listed in 1 year."

"...we have no competitors or partners before, but last year we held out olive branches of the two companies, Sony is one of their quantum is very interested in thin film technology."


"As shown in the image above, Orange represents QuantumFilm absorption capacity, gray are Silicon absorbs light, you can see the gap between the two great. The biggest advantage is that absorption capacity can increase the dynamic range of the camera, dynamic range represents the phone can take a much brighter white and dark black, best embodies quality and filmed under strong light at night. Low light performance is the greatest weakness of the CMOS sensor and mobile phones cannot be one of the main reasons that night."

The interview concludes with a comparison picture with the regular CMOS sensors on the left and Invisage QuantumFilm device on the right. The first row compares full well, I guess, (4000e- vs 12,000e-), the second row is a camera module z-height (5mm vs 4mm), the third row shows the absense of rolling shutter distortions in QF sensors.

11 comments:

  1. Why the power consumption can be cut X8 by using QF please? Jess Lee should be the first Madof in CIS industry.

    ReplyDelete
  2. The diagram is backward I think (although it might just be that I can't read Chinese). The images taken in the dark appear to be swapped. If that is the case, then I applaud their efforts not only for the greater QE at 0.94μm but also in achieving such a feat with pixels nearly 7.5 times smaller than the competitor CMOS in terms of area. I've never been a fan of CMOS.

    Even the James Webb Space Telescope's cameras are utilizing pixel materials other than silicon, which has performed well for many decades in a number of applications, but there's definitely room for improvement.

    How much more do these "quantum film" wafers cost to manufacture in comparison to silicon wafers?

    ReplyDelete
  3. Why has no camera manufacture use this quantum film?

    ReplyDelete
    Replies
    1. The answer can be found in turning this question upside down : the fact that no camera manufacturer has used this quantum film tells you enough about the performance of the device.

      Delete
    2. I can read Chinese. According to this interview, it is said 1.) Invisage had provided samples to some important customers and received very positive feedback. More Chinese end customer can get the samples in July and August. 2.) Camera module makers take the same time (two weeks) as typical silicon sensor to build the cameras. In this case, I think we can very soon know if Invisage and quantum film technology can prove themselves.

      Delete
  4. Maybe the question might be, is there any benefit in using that kind of technology for visible light and modify very expensive and very tunned processes that already deliver high & stable performance?

    ReplyDelete
  5. I remain enthusiastic about any material that has a higher absorption coefficient than silicon, and it would be great in many ways if the QD film is successful.
    On the other hand, most deposited films that have been explored in the past (and it is a long list) suffer from incomplete charge transfer and lag, read noise, dark current issues, ageing issues, and/or PRNU issues.
    Meanwhile single-crystal silicon continues to improve and many of the early advantages promised by Invisage have been eroded by those improvements such as (1) BSI with high QE and high fill-factor, (2) DTI to reduce crosstalk (3) Global shutter development and (4) stacked structure development.
    I suppose the QD film might be picked up by a smaller company first for product differentiation. But, either Invisage offers compelling advantage that causes all the big players to stop what they are doing and switch, or they are not going to be able to deliver on their promises to take over the world. They have been at this a long time now and spent a lot of money. They are trying to chase something that is evolving even faster due to the BILLIONS of dollars of current market revenue and even billions of dollars in investment. It is a hard game to play in. I wish them luck but I am not optimistic. And, I have to say it is a shame they can't just stick to real advantages when promoting themselves, nor deliver on their promises year after year. But, maybe this will be the year?

    ReplyDelete
    Replies
    1. In engineering point of view, yes, it's true. One example I can think of is the voice coil motor which is too mature to allow almost all new motor technologies to enter the camera for mobile phone application. However, I still like to see something which can bring us a different direction, although I do appreciate all the improvements already done on the silicon sensor.

      Delete
    2. The technology differentiation has no value for end-users. Smartphone users would like to see the beautiful image taken with their phone camera despite of the technology used inside. Technology differentiation should be able to provide true advantages compared to the existing one. As well said by Eric, we are facing a running rabbit but not a dead rabbit. My feeling is that every new technology needs a newly enabled application to grow up instead of affronting the existing one.

      -yang ni

      Delete
  6. Seemingly hi-end technology for low-end market. Of course when it comes to prosumer cameras, returns on improving the image sensor provides lesser returns due to other constrains such as optics and cost of lens elements. Think of all those beautiful 500mm lenses with 1:2 or better which will set you back a few grands.

    ReplyDelete
  7. https://www.youtube.com/watch?v=LB4WDj_aBds&list=PLWa6uO3ZUweAZ-VXnnBsDDsBbz32BLlYf&index=181

    ReplyDelete

All comments are moderated to avoid spam and personal attacks.