I would think the signing fee is variable and the nitty gritty detail of their agreements would be commercial in confidence. The contracts would be negotiated depending on no. of nodes used; e.g. Renesas licences 2 nodes. I think there was also a factor of how many different “Devices’ you would use it in; e.g. widget vs phone, vs EV vs smart speaker. I don’t think it is a licence once, put in every device arrangement. You might need multiple licences to put it in multiple different devices.
The price of IP varied depending on what it could do. In this case the uniqueness, first to market, best in class, power saving advantage etc would all work in Brainchip’s favour when negotiating a price. In saying that you could hold out for a top dollar price and end up with royalties of nil because you couldn’t work out a deal that is cost effective for the manufacturer.
I haven’t looked it up and but I recall AM or PVDM during an interview indicating a chip (back when that was an option) was going to be around US $25. A chip would have had all 64 nodes. Therefore 2 nodes would be much cheaper, possibly less than a $1 but I couldn’t begin to put an accurate price on it.
Renesas sells millions and millions of microprocessors and is in the top 5 manufacturers in the world. My recollection is the device including Akida will be available mid year so I would expect royalties to come in after that. It’s not long to wait!
It’s a bit dated but here’s some figures from Renesas:
It’s been discussed numerous times but the contract can be anything lawful both companies agree to. E.g. royalties might not be paid until units are made and sold. But the good news is that once revenue does start coming in it should then continue for many years to come. Over time and multiple customers the revenue flow will build as per Brainchips 2021 conceptual revenue chart from their AGM. They are following their plan.
View attachment 27751
My view is that even if there is no revenue for the next 2 quarters but it kicks in later in the year for me it doesn’t matter. My expectation is that the SP from now will double or triple by the end of the year so 200-300 % is far superior what I would get from the banks!
If for example Qualcomm did use Akida (to complement Prophesee) instead of their current technology in the Snapdragon 8 Gen 2 then that forecast would get blown out of the water!
