1) Microchips don't grow autonomously in the first place, unlike organisms.
2) The article's point is that not only do humans have the "mask artwork and silicon wafers", but also the genes that other organisms use to regenerate portions of their body. A closer equivalence would be "the instructions on how to rebuild the missing half of the microchip from what remains".
Microchips grow autonomously if you regard the entire fab and its staff as a big organism.
And note that the blueprints for the chip do actually encode more or less what it looks like: as a genotype they resemble the phenotype. Yet such a repair is not feasible.
DNA genes do not encode an arm or a leg directly, but a convoluted process by which the structure develops.
Well, we don't really know, because we've never bothered to build a chip/fab process with that capability. It's cheaper to just throw broken ones away (edit: plus, it isn't morally reprehensible to throw them away because they don't have consciousness... yet). I guess the other thing is that a CPU is pretty simplistic compared to a big organism.
We do bin products and then e.g. underclock or disable parts of them accordingly. So there's the start of an "injury"-sensing mechanism. But microchips still have a few thousand years to go in their "evolution" at least :)
The old UK sci-fi television series, Blakes 7, which aired in 1970's had this as one of their predicted technologies.
Entire portions of machinery would regenerate after being exploded in sabotage.
Grids of "radiation mine wire" would after being cut, would have the ends of the cable seek each other out and it would grow to join up again. The strategy to bypass this grid/minefield of interlaced wire, would be to fire the blaster across it clearing a path and then to run through it within 8 seconds before it regenerates.
While it was a bit fantastical for the time, I'm sure some hand-waving could justify nanotech as the way this TV tech might be realised.
