Make Your SoC Upgradable Like A Tesla

I’ve always been a fan of Tesla. Not for the quick acceleration, nice lines, great handling or leading the world away from the using the internal combustion engine. I’m a big fan because they plan products not just for use today, but for the future. In the not too distant past, in order to get the latest automotive technology, you’d have to buy a new car. With Tesla, you don’t have to. You either get an over-the-air update or purchase an upgrade and you instantly have the latest and greatest feature. Tesla owners love this upgradability. Tesla loves it because it builds a devoted user base and additional revenue streams.

Many consumer products already support software upgrades, which are typically focused on bug fixes and plugging security holes. However, the ability to add features via software is limited to the amount of processing headroom available on integrated processors. To add more headroom, chip designers have the choice of using higher performing processors, adding additional processors, or using hardware accelerators. Each comes with pros and cons. Using bigger or more CPUs provides great flexibility but requires more memory and makes the chip architecture more complicated. Accelerators provide great performance improvements but are application focused, so the designer has to gaze into a crystal ball to determine which accelerators to integrate and hope he makes the right choices for a successful product.

There is a third option: Embedded FPGA (eFPGA). eFPGA enables the performance of a hardwired accelerator and the flexibility and programmability like a CPU. There is an area cost for the circuitry to support reconfigurability, but then again, there is an area cost for upgrading or adding CPUs and associated caches. For consumer devices that already use FPGAs, integrating the FPGA on chip will provide system cost savings, offsetting the cost of additional eFPGA silicon with the added bonus of saving power. To keep area cost manageable for consumer applications, we suggest targeting eFPGA to be 5 – 10% of the overall chip area and translates to a 5 – 10% increase in chip ASP.

So, what does a 5 – 10% increase in ASP get you?

• The ability to add functions to consumer devices after purchase that would otherwise require the consumer to buy an upgraded device with a new chip.
• Customer lock-in. Consumers would rather pay a small fee for an upgrade verses paying for a whole new device.
• A potential new revenue stream for selling those new features.
• Targeting more markets to maximize a chip’s ROI. (Different hardware accelerators can be programmed into eFPGA based on the end applications. For example: different XNOR AI networks for a video doorbell vs. baby monitoring vs. retail pedestrian counting).

It’s just a matter of time that consumer companies will follow the Tesla model: You don’t have to buy a new car to get the latest features. Buy an upgrade instead!