I'm reading "smaller, low cost" as microdata (bigger than cube). If you're in LEO you're only up for a few years and totally your use COTs (commercial off the shelf) hardware. How do you deal with radiation? You don't. Hardware fails or your flash it. This is the cheap way to do things. The reason being that these satellites are small and thus launch costs are also small. Here your launch cost is like $100k (which is cheap in the industry). In the small stuff, failure is an option.
When you're launching something bigger you're going into a high orbit (longer lifespan) and you have to deal with more lifetime radiation. Costs are not linear either. So at a larger scale you rad harden because your launch cost is a significant cost to you. We're talking satellites that have tens to hundreds of millions of dollars worth of hardware on them. Not only that, but the satellite took years of work. Here failure is not an option.
But for actual hardening, typically you use sapphire chips instead of silicon. There are other things you can do like encasing in different materials. Multi layered and doped plastics are common now (I actually did some work 3d printing shields, intending to be low cost).
" If you're in LEO you're only up for a few years and totally your use COTs (commercial off the shelf) hardware. How do you deal with radiation? You don't. "
Bitflips even affect systems on the ground. Do you have any data on reliable services using such satellites that ignored radiation in their design? And they kept being offered over time like the companies with satellites that dealt with radiation?
Planet uses COTs last I checked. Most CubeSat companies do (at least that I'm aware of).
I wouldn't say completely ignored though. You still want to use ECC memory when possible and have redundancy. The ignoring that I'm talking about is rad hardware. Or rather sapphire chips. Those run up costs quickly and you're forced to use technology that's on par with silicon from 10 years ago.
When you're launching something bigger you're going into a high orbit (longer lifespan) and you have to deal with more lifetime radiation. Costs are not linear either. So at a larger scale you rad harden because your launch cost is a significant cost to you. We're talking satellites that have tens to hundreds of millions of dollars worth of hardware on them. Not only that, but the satellite took years of work. Here failure is not an option.
But for actual hardening, typically you use sapphire chips instead of silicon. There are other things you can do like encasing in different materials. Multi layered and doped plastics are common now (I actually did some work 3d printing shields, intending to be low cost).