hamachi@KIM.BERKELEY.EDU (Gordon Hamachi) (03/25/86)
For kicks I ran the Brecher disk benchmark on a Mac+ with a RAM disk and with Apple's internal 800K floppy drive. The results suggest that the disk cache makes little difference. Is that right? Also, the RAM disk is blazingly fast, and no hard drive will come close to its performance. >The benchmark consists of three parts: >(1) 100 reads of 32KB of data from the start of the volume; >(2) 100 writes of 32KB of data to the start of the volume; >(3) 40 iterations of: read one 512-byte block from an offset of 1MB, followed > by read of one 512-byte block from start of volume; >( times are in ticks, i.e., sixtieths of a second): > > Data transfer time Access Time > ------------------ > Reads Writes > Apple Internal 800K drive (1) 8719 11543 (5) Apple Internal 800K drive (2) 8324 11540 (5) >Tecmar MacDrive (3) 6017 6719 401 >DataFrame 20 1344 2233 487 >HyperDrive (4) 1586 1600 563 >MicahDrive 20 AT 507 508 528 400K RAM Disk 192 192 (5) > Notes: (1) Disk cache disabled (2) 96K disk cache (3) 10MB Fixed (4) original 10MB unit with old controller and version 5.1 software (5) Volume to small to complete this test .
ephraim@wang.UUCP (pri=8 Ephraim Vishniac x76659 ms1459) (03/27/86)
> For kicks I ran the Brecher disk benchmark on a Mac+ with a RAM disk and with > Apple's internal 800K floppy drive. > > The results suggest that the disk cache makes little difference. Is that > right? I'm surprised that the cache makes any difference at all. The benchmark program makes direct calls to driver (so the posted description says; I haven't looked at the code). Reasonably, this should bypass the Apple disk cache code. The caching should only affect file I/O, not direct disk I/O.