Microchip Production 101 for Klaus
Correct. Not every wafer yields the same number of chips. Of course they tweak and tweak and tweak until they get everything setup so they get optimum usable chips. Then, they get new raw material, so they tweak and tweak, and so the process continues.
Even if they do produce good chips, that is, the circuits are whole; nothing broken, these chips still have to be tested. There are normally failures here too. So, hypothetically speaking, after all the tweaking, they get say 85% yield, and then anywhere from 5-10% failures. That's why IBM uses such a large wafer. Makes it more cost effective. It makes processors too damn expensive when you are producing smaller yields on smaller wafers (150-200mm)
klaus said:Correct? Cause I don't follow the manufacturing of processors very closely to know all this **** 🙂
Correct. Not every wafer yields the same number of chips. Of course they tweak and tweak and tweak until they get everything setup so they get optimum usable chips. Then, they get new raw material, so they tweak and tweak, and so the process continues.
Even if they do produce good chips, that is, the circuits are whole; nothing broken, these chips still have to be tested. There are normally failures here too. So, hypothetically speaking, after all the tweaking, they get say 85% yield, and then anywhere from 5-10% failures. That's why IBM uses such a large wafer. Makes it more cost effective. It makes processors too damn expensive when you are producing smaller yields on smaller wafers (150-200mm)