Thought I would share some thoughts on the revenue from nandox should we finally get that EUA.
A production cost model based on available information looks like this:
- Our production capacity is likely to be a little greater than 1000 150mm wafers per month, given we have been trying to expand capacity, and historical capacity was given as 1000 per month (https://en.wikipedia.org/wiki/List_of_semiconductor_fabrication_plants). I'll call it at 1300 per month (a 30% recent increase)
- Cost per wafer from the Lawes' cost model previously posted comes in at $1040 per wafer, but I'll call it $1100 to add a bit of buffer
- Our gross margin is at 57%, so I'll leave that unchanged
- From pictures, our chip size is roughly 5mm x 5mm
- I'll assume a 90% yield which is what should be expected from an optimised production line
These parameters (all of which could be tweaked), result in revenue per chip of $4.92.
With a 3 month full production run, this results in US$10M in revenue. So, I'm pretty much convinced now that the $10M figure will turn out to be our quarterly revenue from our direct production of the sensor. This depends on the above parameters being roughly right.
The 3 month run only results in a little over 2M sensors - clearly no where near enough to satisfy the likely demand for the nanodx COVID sensor.
In the AGM presentation, Ralph indicated we were looking for strategic partnerships to add scale/capacity by about mid this year. This kind of strategic partnership in the MEMS Fabrication industry looks like this, a description taken from a MEMS product development book:
So, it's not uncommon for the foundry that develops the sensor to the point of production (Sensera) would then licence production to a larger Foundry that has the scale to met demand.
Let's consider what that might look like. Skywater, which Biden visited, and its Execs have liked Nanodx posts, operates at a gross margin of around 20% from its latest report. Skywater runs 200mm wafers, so it's production costs will be lower than Sensera's - by about 20% from the Lawes model.
This means that Skywater would meet it's standard gross margin (cost of chip at $1.69, revenue of $2.11) with Sensera taking up to a whopping $2.81 clip as the royalty, with the same sale price to Nanodx ($4.92). The clip would be unlikely to be this big for other reasons no doubt I guess. The point is that there is enough head room for Sensera to earn a substantial return from licencing to a Fab like Skywater.
So, here is the big speculation. Given Skywater's interest in Nanodx posts, could Ralph have landed a royalty agreement with Skywater?
A royalty of US$1 per chip to Sensera would create an enormous revenue stream - all cream, no cost. It would easily outstrip the revenue Sensera would generate from direct production of 2M chips.
Overall:
- Direct yearly revenue of US $40M, with $22.8M gross profit
- Indirect royalty revenue (assume $1 royalty, with 50M chips, from only about 4000 wafers per month for Skywater) - $50M gross profit.
Now, there is something to raise a glass to on a sunny saturday afternoon.
