Appeal No. 95-4899 Application 08/131,029 look to Harris for a teaching of a FIFO buffer to add to Moreno. After all, Moreno never discusses a “peripheral data flow problem” (Answer, page 4), and we are not aware of such a problem in Moreno. Even if we assume for the sake of argument that it would have been obvious to one of ordinary skill in the art to add a FIFO in Moreno, we are not convinced by the examiner’s conclusion (Answer, page 4) that “it would have been manifestly obvious to . . . incorporate a FIFO buffer in between the cache memory and the disk of Moreno’s system” (emphasis added). Other than appellant’s disclosed and claimed invention, nothing in the record would have suggested such a specific location for the buffer. In summary, the obviousness rejection of claims 14, 23 through 25 and 27 through 30 is reversed because “nothing in the prior art of record suggests incorporating a buffer between a mass storage device and a cache memory” (Brief, page 15). Ryan discloses a dual or two-part cache memory (Figure 2). A first cache memory 20 handles instruction data, and a second cache memory 22 handles operand data. The data flow through one cache memory is independent of the data flow 5Page: Previous 1 2 3 4 5 6 7 8 9 NextLast modified: November 3, 2007