As alwaysl, Doni, you are right on target with this. I've been going back to re-read the original Markman ruling, which thanks to LGJ can be found here. In doing so, I focused on Item F. “a path for sequentially accessing the data segments within the primary memory."
For sure, this is a complex issue, but this section makes clear that while '108 doesn't forbid the use of FATs stored in volatile RAM, neither does it require them and that its definition of "primary memory" doesn't require RAM. (See below)
Specifically, Micron points to the patentee’s statement that Ban, Lasker, and Jeffrey “try to force a FAT on flash memory which the present invention realized was totally inefficient.” (Id. at 19; Doc. No. 43-4 at 8.) Further, Micron cites the patentee’s statement that “[t]he applicant does not understand why the concept of using a FAT is being asserted here as making the present invention obvious when the FAT has been eliminated from the present invention.”
...........................................................
In that prior art, each time a file is to be modified, the file data must be retrieved from flash memory, placed into a volatile RAM cache, manipulated in RAM, and the entire file resaved to flash memory in a new and unused contiguous memory segment. (See ’445 File History, November 10, 1997 Amendment at 7 (“Ban retrieves an entire file which is to be modified into volatile RAM cache memory, manipulates the data ..., then resaves the entire file to main memory in a new and unused contiguous memory segment”); see also id. (“It should be apparent from the description of Ban in the patent application that it uses flash memory just as the combination of Lasker and Jeffrey would try to use it.”); see also id. at 6 (referring to the invention of the ’445 patent: “instead of reading a FAT into cache memory, erasing the FAT from primary memory, modifying the FAT in cache memory, and then resaving it to primary memory each time that any data is added to or deleted from primary memory (as Ban does), no FAT is used, and data is manipulated directly in primary memory").
..........................................
Although portions of the specification refer to FAT generally, the inventors only discussed and distinguished a specific feature of FAT. Further, the ’445 patent’s references to embodiments using a memory block map, root directory, the “Norris File SystemDirectories,” “index of segments,” and “presentation map” demonstrate that the ’445 patent’s distinction of certain prior art memory mapping technologies does not amount to a “clear and unmistakable” disavowal of memory mapping or file allocation tables in any form. See 3M Innovative Props., 725 F.3d at 1325; Omega Eng’g, Inc. v. Raytek Corp., 334 F.3d 1314, 1325 (Fed. Cir. 2003).
..............................................
So, as I understand it (JMHO), '108 offers a broad definition of Primary Memory that has a plain and ordinary meaning that may or may not include RAM. It also allows for the use of FATs but does not require them. But what it does do that is unique and original is manipulate the data directly in Primary Memory without using a Cache or requiring FATs or dynamic RAM, a process is much more versatile and efficient than prior art. And is also broader and more inclusive.
