Its not quite that simple. So I decide I am going to start making memory. I do all my up front capital investment. Now I have to decide how much of my fixed costs I want to try to recoup per unit. One question I might ask myself in the chip industry is how long will this stuff be in mass market demand. Nobody will want my chips if a new tech comes out that doubles density. My current equipment won't be useful anymore. Now I don't know when this will happen so I am going to probably start off with higher prices, pessimistically assuming the window of viability will be small. My competitors are naturally doing this as well so we can all charge high prices.
Suppose a few years have gone by and there have NOT been any major process improvements. My initial capital investment is paid off. My variable costs have been controlled as well as they can. My contribution margin is maximized fully. I now have every incentive to sell as many units as possible! So its a question of capacity. If can produce 20% more chips running another shift or something I probably will. If I have to cut my prices to sell those chips some, I probably will still do it. On the other hand if I don't have spare capacity, I probably don't want to expand my capital investment into three year old technology. My competition may or may not be in the same position. If all of us are selling all the chips we can produce at current prices, than nobody has any reason to lower prices.
The moment it looks like a new technology is coming down the pike, even if its just a die shrink though suddenly we have inventory to clear..