amorsin@cs.vu.nl (A.W.Morsink) (05/06/91)
Does anyone know if it's possible to use synchronous transfers (SCSI) on the CBM A2091 controller? It is supported by the Quantum ProDrive HD, rite? I ask this because I recently came across a utility called "battmem" that enables A3000 users to "initiate syncronous SCSI" by flipping a bit in it's battery back-upped RAM. How much difference with the 3000 design is there? Is it just a software pro- blem that could be fixed? I am planning to buy an A2630 board, how much of the transfer rate that can be attained is a result of sync. vs async transfers (all limited by the ZorroII bus speed since it can't 32-bit-DMA into the A2630's memory)? Thanks, Marco Niese (Using: amorsin@swi.psy.uva.nl)
daveh@cbmvax.commodore.com (Dave Haynie) (05/08/91)
In article <9854@star.cs.vu.nl> amorsin@cs.vu.nl (A.W.Morsink) writes: >Does anyone know if it's possible to use synchronous transfers (SCSI) on the >CBM A2091 controller? It is supported by the Quantum ProDrive HD, rite? I ask >this because I recently came across a utility called "battmem" that enables >A3000 users to "initiate syncronous SCSI" by flipping a bit in it's battery >back-upped RAM. >How much difference with the 3000 design is there? Is it just a software pro- >blem that could be fixed? I don't think there's any way to automatically have the A2091 initiate synchronous SCSI, but you can certainly have it done "manually", via some kind of program in the Startup-Sequence. At least I think so; the A2091 and the A3000 use different DMA chips, but the same SCSI chip. The A2091 may have a slower clock used for the SCSI chip, I'm not sure about that. >I am planning to buy an A2630 board, how much of the transfer rate that can be >attained is a result of sync. vs async transfers (all limited by the ZorroII >bus speed since it can't 32-bit-DMA into the A2630's memory)? The Zorro II bus speed limits transfers to roughly 3.5MB/s. While synchronous SCSI can theoretically hit 5MB/s, the controller determines the actual speed. Most of the time the physical limit is the actual hard disk mechanism, not the speed of either the Amiga bus or the SCSI bus. The 32 bit DMA of the A3000 won't do much to increase the speed of a single SCSI drive. It would let you run some kind of multiple disk scheme, or anything else that can result in sustained SCSI bus peak speed activity, faster than on an A2500-class system. Most of the time, the A3000's design just results in more available CPU time during the same SCSI operations. >Marco Niese -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "That's me in the corner, that's me in the spotlight" -R.E.M.