Marketwired: Horizon Technology Finance Corporation, a finance company that provides secured loans to venture capital and private equity backed development-stage companies, announces it led a venture loan facility, in which Square 1 Bank participated, for InVisage Technologies, Inc. InVisage will use the funds primarily for working capital purposes.
"We are pleased to add InVisage to our top tier list of venture loan portfolio companies," stated Gerald A. Michaud, President of Horizon. "InVisage's QuantumFilm platform and series of products address the fundamental physics limitations of today's silicon-based image sensors used in smartphone cameras and other applications. This loan facility provides InVisage with strategic capital to support the manufacturing and commercialization of its products."
InVisage President and CEO, Jess Lee, shared, "We are pleased to have Horizon as one of our investment partners and are fortunate to have attracted top-tier investors that share our vision of creating the next era of cameras -- fast, thin, high performance -- to transform the mobile, photography and connected device segments. This significant loan facility is an affirmation of the growth opportunity we see for our QuantumFilm™ platform and series of products. We now have additional financial flexibility to drive and market our capabilities to Tier 1 customers."
In April 2014, Invisage announced its latest funding round of $18M, bringing the total investment to more than $100M.
Need more money = not enough revenue
ReplyDeleteHow would a quantum film make cameras faster and thinner?
ReplyDeleteSpeed is dictated by the CMOS readout, and module height is determined by the lens TTL + silicon thickness + substrate thickness. I don't see how quantum films affect any of those outside make-believe land of marketing speak.
I guess "speed", here, is about the possibly increased QE that allows you to capture images at increased frame-rate - given a fixed, desired SNR. "Thinner" is probably about the different absorption mechanism as well. In a normal imager you need a certain epi-thickness caused by the comparatively poor absorption coefficient of silicon. With the approach Invisage wants to use, you yield a much larger absorption coefficient, resulting in a thin photo-detection layer on top of your pixel, what eliminates the necessity of a thick epitaxial layer and also results in an improved fill factor. Thus, you may ultimately end up with a thinner device. The actual disadvantage of this approach that still seems to be solved is the excessive dark current these dedicated absorption layers imply. Concerning your doubts, however, I think what they claim is meaningful - it's not all just marketing ;-D
Delete- Andreas Süss
If the dark-current is higher as you state, then you will not probably achieve the desired SNR even though you may have higher QE. The thickness of a state-of-the art BSI sensor would be around 3.5 um (min). You need a carrier to handle the thin EPI. Since you need a carrier of sufficient thickness for reliability, making it thinner does not make any sense. Finally, the optics decides the final thickness of the device. A few microns doesn't.
DeleteA thinner EPI does not result in a larger fill-factor. The amount of metal blocking layers decide that. BSI have 100% fill factor.
Concerning the fill-factor - that's what I meant. I suppose Invisage, here, compares against FSI. Their approach also achieves 100 % fill factor as the absorption layer is on top of the metal stack. Concerning the dark-current, I agree - dependent on integration time it might be difficult to achieve similar SNR. However, for high-speed cameras with short integration times I think the larger dark current won't be much of an issue and you probably can actually end up with higher frame rates. Concerning the thickness - indeed you need a handle wafer. So, I agree that this is a more marketing biased claim. However, having a 3.5 um Si Epi layer certainly isn't advantageous for red towards infrared light where equivalent penetration depth is roughly 8-10 um. If you want to end up with a high QE in that wavelength range, you need a thicker epi if you stick with Si and thus possibly get crosstalk issues for small pixel pitches if you don't take further actions. I think the Invisage approach has advantages here. Nevertheless, I agree, ultimately the overall device won't be thinner. So, basically I think that the advertisement Invisage is performing may sound a bit general - but if you think about, you can certainly find applications were (most of) their claims actually make sense. (by the way - I am not working for or with Invisage..., so I'm not trying to take sides, here)
Delete- Andreas Süss
I had a look at their site. They claim that when you decrease the camera lens height, color xtalk gets worse, due to the distance from color filter to the photodiode. Since in their case the CFA and photodiode/sensitive element are very close, it allows them to reduce the camera lens height while keeping reasonable color xtalk.
DeleteThe image they show is funny, though. They compare it with a FSI sensor, and the sensor metal stack is almost as high as the camera lens focal length. So it makes it look like just marketing.
Andreas,
DeleteThe die thickness has little to do with the epi thickness and everything to do with the mechanics of handling thin silicon. In a BSI CMOS sensor the epi is only a few microns thick, but the typical die thickness is 200µm with most being the handling wafer.
If those few microns of epi were replaced with even a zero thickness quantum film, the die would still be back ground to 200µm for mechanical reasons.
There is some increase in crosstalk due to high CRAs even in BSI (their showing of an FSI pixel in Figure 1 in the Technology section of their website is disingenuous), but there are methods to improve that which are under development and don't require any new wonder material.
DeleteIn any case, we're unlikely to see any significant reductions in back focal length and corresponding for planar silicon in the foreseeable future due to the constraints it places on the lens design and the more significant color shading factor that's caused by the IRCF angular shift.
This claim of thinness is just marketing. There's little practical benefit.
All talk, no game.. I'm waiting years to finally see any output from this "ground-breaking" tech.. not that i'm a hater but I'm getting tired of all the marketing rap
ReplyDeleteI am looking forward to Invisage quantum film used with a Pelican Imaging array camera, then put inside a Leap Motion device. It will be a critical mass of marketing hype, perhaps creating a rupture in the time-space continuum.
ReplyDeleteI would not put LeapMotion in the same category. LeapMotion has made a dreaming demo video, but it is a real product sold in many shops in the world. Their product can please or not please users, but all is real and working.
DeleteQE for CMOS sensors was low when Invisage started but has become so high now that there's little room for improvement.
ReplyDeleteSimple question: what means "secured loan" please??
ReplyDelete"Secured loan" means that the money was loaned against hard assets used as collateral. The assets could be physical, like buildings and equipment or financial, like bonds.
Delete"Unsecured loan" means you got the money because the bank likes your smile.
Thanks Dave!! This means that this is their last bullet.
ReplyDelete