Forum unknown
· Hardware
#hard disk installation
9 messages in this thread
Question: At what point does AmigaDos becoming the limiting speed factor in
hard drives? I would be most interested to know.
At 137,900 bytes per second.
This is the speed at which Facc peaks (as measured by DiskPerf). As Facc
uses the standard file system handler and the RAM: device which uses its
own handler (and reaches speeds of 800,000 bytes per second using the same
test) we can infer that the standard file system's maximum through put is
the limiting factor.
Truely high performance will be acheived through the use of alternate file
system handlers or through the use of co-processors implementing greater
and greater sections of the disk subsystem directly on card.
As for the ASDG SDP, you can expect a stage 1 SDP to be very near the
137,900 bytes per second. Stage 2 will be higher. Stage 3, higher still.
These stages represent succesive generations of SDP software (same
hardware).
Does this help to answer your question?
Perry S. Kivolowitz – ASDG Incorporated – (201) 563-0529
My 8 Mhz AT clone, 30 meg, 30 ms drive, 90 percent full, registers 165-170K
per second throughput on the latest Core International hard disk test. Mac
drives have higher performance than that in burst modes, as I recall. We're
still behind.
My 8 Mhz AT clone, 30 meg, 30 ms drive, 90 percent full, registers 165-170K
per second throughput on the latest Core International hard disk test. Mac
drives have higher performance than that in burst modes, as I recall. We're
still behind.
Think so. Thanks. Am I to infer that you intend to bypass AmigaDos in
your future generations of software to gain additional speed, since you
mention using the same hardware and you are currently running near the
limitations of the existing drivers?
Yes. To understand the extent of the by-pass, recall that the SDP has an
on-card 68000, 512K of ram, and its own MMU (as well as some roms for
getting things going). In essence – nothing that is disk related will even
run on the Amiga (except for your tasks, that is) once we reach level three
of the SDP software.
Perry
Yes. To understand the extent of the by-pass, recall that the SDP has an
on-card 68000, 512K of ram, and its own MMU (as well as some roms for
getting things going). In essence – nothing that is disk related will even
run on the Amiga (except for your tasks, that is) once we reach level three
of the SDP software.
Perry
Think so. Thanks. Am I to infer that you intend to bypass AmigaDos in
your future generations of software to gain additional speed, since you
mention using the same hardware and you are currently running near the
limitations of the existing drivers?
At 137,900 bytes per second.
This is the speed at which Facc peaks (as measured by DiskPerf). As Facc
uses the standard file system handler and the RAM: device which uses its
own handler (and reaches speeds of 800,000 bytes per second using the same
test) we can infer that the standard file system's maximum through put is
the limiting factor.
Truely high performance will be acheived through the use of alternate file
system handlers or through the use of co-processors implementing greater
and greater sections of the disk subsystem directly on card.
As for the ASDG SDP, you can expect a stage 1 SDP to be very near the
137,900 bytes per second. Stage 2 will be higher. Stage 3, higher still.
These stages represent succesive generations of SDP software (same
hardware).
Does this help to answer your question?
Perry S. Kivolowitz – ASDG Incorporated – (201) 563-0529