Who Will Make Nikon's Next Sensor?

The recent Nikon revelations have started the old "Sony makes all Nikon sensors" conspiracy theories all over again. Just a reminder, those statements have never been accurate. In the last 20 years Nikon has used image sensors manufactured by Aptina, Renesas, Toshiba, and Tower Semiconductor, as well as Sony Semiconductor. 

What do we know? (a) That the upcoming Nikon Z9 flagship camera will feature a stacked sensor (so it will likely be BSI also). (b) Nikon also just revealed an image sensor research project that involves a 1" BSI stacked sensor prototype with both the image sensor and the companion semiconductor fabbed at the 65nm process size.

So, some other facts: at the moment I'm aware of three companies that are delivering production quantity, stacked image sensors (mostly at smartphone sizes, though): OmniVision, Samsung, and Sony Semiconductor. A couple of others, including Nikon, have now demonstrated R&D projects with stacked semiconductor technologies. 

Neither OmniVision nor Samsung seem to be making large format sensors at the moment, so the natural assumption is that Sony Semiconductor would be the only one able to provide Nikon a stacked image sensor. Sony's known stacked sensors, however, appear to utilize two different-sized process sizes bonded together, and I don't believe either of them are 65nm. Thus, this probably means Nikon's 1" research project was fabbed at Tower Semiconductor, which began a 65nm image sensor program in Japan back in 2014. In other words, we have another possible large sensor producer that now has stacked capabilities, and on a fab local to the Nikon research team.

Tower in Japan currently processes 12" wafers with 65nm BSI process and copper wiring, and supports stitching, dual light piping, and global shutter, pretty much all things we might expect in a state-of-the-art full frame image sensor. And yes, the 2.7 micron pixel size Nikon announced for that 1" sensor project also fits on that fab (it can go as small as 2.5 micron with global shutter), and that's smaller than we need for a new full frame flagship camera. 

Thing is, Nikon makes the equipment (steppers) that makes semiconductors such as image sensors. While ASML has the largest market share for steppers by far, it is currently mostly catering to companies like Qualcomm, Samsung, and TMSC outside of Japan. In Japan Nikon (and Canon) have done far better selling steppers to the local electronics companies. That's how Nikon's sensor group ended up at times during the digital age working with Renesas and Toshiba for image sensor supply. It's probably how the Sony/Nikon relationship first started, though it's unclear just how good that relationship is today. 

So, I'd conclude that there are two likely possibilities for Nikon's flagship camera image sensor: that they somehow got Sony Imaging—remember that group claimed that they paid for much of the R&D work on the A9 sensor, which was the first full frame stacked sensor—to agree to let them use (a derivative of) the existing sensors, or that Nikon specified a new sensor to use and/or created their own new technology. In that latter case, it very well could be Tower Semiconductor that fabs Nikon's upcoming sensor, which would almost certainly mean a new Nikon-specified design. That could also explain why the flagship camera is arriving last in Nikon's Z full frame lineup: full sensor development has very long lead times. 

Let me be clear: I don't know which is going to be the case for Nikon's Z9 flagship (Sony or Tower), but I'm nearly certain one of them will turn out to be the supplier. My reading of the tea leaves seems to suggest it is more likely Tower than Sony. Others will claim it is more likely to be Sony than Tower. None of us know.

Do I care who makes the sensor for Nikon's upcoming flagship camera? Nope, not at all. Nikon has shown good judgment virtually throughout the digital era in their selection of technologies and sensors to use. Even some of the maligned cameras—the D2h comes to mind—actually had very good image sensors in them for the time. The D2h's "flaws" were in the lack of adequate near IR filtration and only having 4mp when Canon was already double that, not particularly the image quality, which I found quite good. All of the current Nikon cameras, DSLR and mirrorless, have image sensors that can claim to be at or near state-of-the-art. 

Unfortunately the secretive nature of Nikon's engineering coupled with the insistent demand of the Internet to speculate in lieu of actual information means that we're once again getting the old "Nikon is stuck with Sony" type of posts. I don't believe that's ever been true, and I don't believe it to be true today. I've outlined a few things that are known above, but that actually wouldn't completely rule out a third option: that Nikon will use a sensor from a source not yet known. 

Again, I don't care whose name is on the back of the chip where it will never be seen (and etched in code in the silicon where you'd need special equipment to see it). What matters is how the sensor performs, which is obviously not going to be known until a camera appears and we can all test it. Which is still a ways off (sometime in 2021, likely late fall).

In the meantime, I'm enjoying the current Z's, including the Z6 II and Z7 II, which represent Nikon's best work so far in mirrorless. That we have confirmation that there will be a better camera soon is nice, and I can't wait to see what Nikon has been up to. I'll bet that most of you will still be buying a Z6 II or Z7 II though ;~).

Looking for other photographic information? Check out our other Web sites:
DSLRS: dslrbodies.com | mirrorless: sansmirror.com | general/technique: bythom.com | film SLR: filmbodies.com

text and images © 2021 Thom Hogan — All Rights Reserved
Follow us on Twitter: @bythom, hashtags #bythom, #zsystemuser