<< Curious, do memory manufacturers use the same or similar manufacturing processes that Intel and AMD use (lithography), and if so at what resolution (in microns)? With the recent spike in memory pricing, I am wondering if this is more gouging by the manufacturers then a real hardship cost-of-development case that are simple passing the costs down the line. >>
Yes it would be safe to say that the method by which memory is manufactured is very similar to that of processors. Memory makers have the advantage of not having to use as many levels of metal interconnect and thus lithography steps are reduced. They also have an advantage over Intel and AMD in that their products are not as sensitive to process defects that in a processor could make that 1GHz Athlon run at 700MHz and lose the company a lot of money. Their products are always clocked relatively low. Memory is also fairly simple to produce as it is fairly uniform pattern accross a large area (relative). This helps to aid in manufacturing process such as CMP (chemical mechanical polishing) and etch where you realize a bias between dense and isolated features. In truth memory makers have historically lead the advancement of new lithography technologies in an ongoing effort to reduce the size of their memory cells. (Moore's law and such is usually measured by the DRAM industry). You see since the design of memory really never changes it is simply a function of how many chips they are able to physically fit onto a single wafer. Thus they benefit greatly from increased substrate size and smaller lithography features. Some manufacturer's are currently working on .10 and .11 micron technology and most will move to .13 before the larger logic manufacturers like Intel and AMD have completed their transitions. Another advantage memory manufacturer's have is the sheer volume of product they produce over which to amoritize their costs such as photomasks and capital equipment. They produce a lot more product on the same tools than a company that is forced to change designs rapidly or with low volume.
Disadvantages of the memory industry include the fact that they are highly sensitive to any particle defects since a particle can render an entire cell as unusable. Also they operate at relatively low margins since cost is a function of silicon needed per die and for the money they receive they use a lot of silicon. But as you say their R&D overhead is greatly reduced unless they are keen on developing new memory technologies for the future. Also depending on the type of memory (i've been speaking mostly about DRAM) and the method of production they might come up against challenges not present in the semiconductor world. For instance in order to fit the amount of memory they do in the space they do the transistors are very tightly packed. One way of achieving very high density is to build the transistor devices vertically into the silicon rather than using a very thin layer at the top like CMOS logic. Thus they need a greater portion of their wafers denuded of defects to the silicon lattice such as oxygen precipitates, interstitials and dislocations/stacking faults. This can drive the price of their starting material up.
Since the majority of DRAM manufacturers lost large amounts of money last year I would have to doubt that they are price gouging you. They are forced into the difficult position of trying to juggle supply and demand.