InTheSand wrote:
I'm surprised! I was under the impression that the IDE connector on the Buddha was just as slow as the A4000/A1200's internal connector. But I'm willing to be corrected!
- Ali
The internal A4000 IDE connector is pants. To quote the AHD:
The built in IDE controller supports two IDE devices connected simultaneously. Unlike the A1200 the A4000 has a buffered IDE port. Only mode PIO 0 is supported.
whereas:
Fast EIDE controller
supports PIO0, PIO3 and PIO4 devices
meets the ATA 3 and Fast ATA 2 specifications
up to 16.6 MB/s transfer speed
two 40 pin IDE headers (primary and secondary)
the primary and secondary buses can be accessed at different speeds
up to four IDE or ATAPI devices can be connected at once
hard disk activity LED connector
unconventional handling of >4 GB devices, they are simply split into separate logical 4 GB blocks - can be turned off by software for filesystems implementing NSD, TD64 and Direct SCSI commands
multiple FastATAs are supported
IDE controller
uses polled I/O, not DMA transfer
two buffered IDE ports support up to four IDE devices
each port is compatible with IDE splitters allowing up to a maximum of eight drives
autoboot ROM
two LED port activity connectors - one for drives 0 to 3, the other for drives 4 to 7
software configurable IDE timing - even PIO mode 0 devices are compatible
raw transfer speed is limited by the Zorro II bus to 3.58 MB/s
supports hard disks larger than 4 GB
can mount GVP or AT-Apollo formatted hard disks
supported by Linux
Sysinfo certain reported a huge increase in speed with the FastATA over the internal IDE header.
For example, if I set the FastATA to PIO 0 as used by the internal header I get a speed value of 1,191,563 bytes per sec.
But, when I use the FastATA in PIO 4 mode it jumps to 6,553,600 bytes per sec.
PIO 3 = 3,276,800 bytes per sec
Tests on a Western Digital EIDE WD800 3.5" drive.