- ATA is the lowest priority, but the ATA port of VIA6421A seem to work.
- SATA for MacOS "9" is fine on Silicon Image 3112, 3512, 3114.
I assume, it is fine on Vitesse 7174 (alias SeriTek/1V4), but since that chip is very rare and nearly unheard on anything, but FirmTek - the priority is relatively low (FirmTek is OK, but needs to be updated).
The insides of the Vitesse 7174 are somewhat similar to "Frodo" (the Broadcom 5770) which is more common.
The Frodo has either 8 or 4 ports. Almost all Frodo controllers I seen are based on BCM5770 8-port chip, not on BCM5770R 4-port chip.
In addition, Frodo made it inside of the Apple G5 SATA.
But: do not rely on Apple code (otherwise open source as "K2 chip"). And, of course, there is no "9" driver from Apple, only "X".
And Apple's code does not work to jump-start most of SSD-s, this is why most SSD-s don't work in G5.
That's not G5 fault, not SSD fault - but fault of Apple's OF and "X" driver for Frodo.
I have bunch of Frodo controllers (you can find these under RaidCore 4000 name on fee-bay). Without developer's documentation, just based on Linux / FreeBSD / Darwin code you can forget about driver writing. Without Darwin, Linux, FreeBSD (just having the Broadcom documentation for developers) you can forget about driver writing. Nasty.
Currently at 6.0.0 "Beta 22" and still suffering like a dog from that "Lord of Rings".
Otherwise all Silicon Image first generation SATA and VIA6421A things look good enough to become public "beta".
SAS: first experiments with forcing the 3.3 Volt LSI SAS PCI-X controller into 5 Volt PCI-X slots of G4 look pretty good.
To bad, the home-made isolation made in a Bavarian village two days ago broke, so for now not much I can do.
Otherwise the said isolation + a proper riser card should make the card appear in the system and not shot down the p/s
- SATA for MacOS "9" is fine on Silicon Image 3112, 3512, 3114.
I assume, it is fine on Vitesse 7174 (alias SeriTek/1V4), but since that chip is very rare and nearly unheard on anything, but FirmTek - the priority is relatively low (FirmTek is OK, but needs to be updated).
The insides of the Vitesse 7174 are somewhat similar to "Frodo" (the Broadcom 5770) which is more common.
The Frodo has either 8 or 4 ports. Almost all Frodo controllers I seen are based on BCM5770 8-port chip, not on BCM5770R 4-port chip.
In addition, Frodo made it inside of the Apple G5 SATA.
But: do not rely on Apple code (otherwise open source as "K2 chip"). And, of course, there is no "9" driver from Apple, only "X".
And Apple's code does not work to jump-start most of SSD-s, this is why most SSD-s don't work in G5.
That's not G5 fault, not SSD fault - but fault of Apple's OF and "X" driver for Frodo.
I have bunch of Frodo controllers (you can find these under RaidCore 4000 name on fee-bay). Without developer's documentation, just based on Linux / FreeBSD / Darwin code you can forget about driver writing. Without Darwin, Linux, FreeBSD (just having the Broadcom documentation for developers) you can forget about driver writing. Nasty.
Currently at 6.0.0 "Beta 22" and still suffering like a dog from that "Lord of Rings".
Otherwise all Silicon Image first generation SATA and VIA6421A things look good enough to become public "beta".
SAS: first experiments with forcing the 3.3 Volt LSI SAS PCI-X controller into 5 Volt PCI-X slots of G4 look pretty good.
To bad, the home-made isolation made in a Bavarian village two days ago broke, so for now not much I can do.
Otherwise the said isolation + a proper riser card should make the card appear in the system and not shot down the p/s
Last edited:
