Well, in a manufacturing environment, everything is in cleanroom. Manufacture, processing, testing, packaging, all of it. All in varying degrees of cleanliness, but all in cleanrooms. Manufacture and processing are the cleanest, because those are the ones where real damage can be done (a dust particle in the middle of a half-finished chip is worse than a dist particle sitting on a finished chip).
The "class" of a cleanroom tells the level of "cleanliness", as it relates to airborne particulates (which is what that suits are for, people shed a TON, by far the dirtiest things in any given cleanroon). The number of the class refers to the average number of >500nm particles per cubic foot of air. Eg, Class 10 cleanroom has less than 10 particles of a size greater than 500nm per cubic foot of volume in the room. If you've ever seen dust in the air in bright light, you know it's a lot higher than that typically, and 500nm is smaller than the smallest dust that you can see with naked eye (typically around 1000nm is that limit). So that tells you how clean it is.
So chips (referring to silicon computer chips) and electronic devices like diodes are manufactured and processed in really clean environments, and every stop after that gets progressively less clean. Probing is still pretty clean, since bare die aren't packaged yet (testing a device while it's still on a wafer), packaging is fairly clean (this is where they package the devices, chips into those black cases that are what you see, diodes into cans, etc.), and testing is relatively not very clean, since things are all nicely sealed up. In an industrial test lab, for instance, you might see just a lab coat, a hair net, and grounded shoe covers to prevent static, all this as opposed to a full bunny suit in a manufacturing setting down the hall.
In university research, it changes a lot. Money isn't riding on it as much, and there isn't as much money available, so everything is an order of magnitude less clean. Diode crystal growth is done inside chambers, so it's pretty clean; but the systems aren't in cleanrooms, so it's dirty going into and out of the system and stored in a clean box. Processing to make a device is done in a cleanroom typically, but not as clean as industry. Testing could be anywhere. We have optical testing scattered everywhere, with some in cleanroom and some in normal lab space on airtables.
Basically, it really just comes down to how much money and resources you have and how much is riding on the stuff you're making. Industrial manufacturers will spare few expenses, where universities will get by doing the best they can with what they have.
