The card itself is a half height half length server sized card using the PCI Express x4 interface to communicate with the motherboard. This baby will fit into a x4, x8 or x16 slot on any PCI-E based motherboard. You can put it into a x2 or x1 universal slot, but only if its an open ended slot. The front of the card houses the SATA ports, controller chips and the LED jumpers.
The back of the card is rather bare. Only the BIOS chip as well as the Highpoint HPT601 hot swap control chip is located here.
To power the card, Highpoint has chosen to uses its rival, Marvell’s 88SX6081-BCZ controller chip. This chip is the first to support PCI Express on a large SATA scale. The chip support up to eight SATA-II or SATA-I devices on a single chip with the ability to span link two cards together to give a total of 16 drives in a single array, very large redundancies available here. The chip also supports RAID 0, 1, 5, 10 and JBOD functions, though why JBOD or RAID 0 would be used in a server environment is beyond me, but the features are there.
You will notice a rather large heatsink just to the left of the unit. This covers an Intel SAF-TE controller chip. SAF-TE or SCSI Addressed Fault Tolerance Enclosure is a specification of defined SCSI commands for setting drive status. This allows the controller to monitor the status of the drive’s RPM, SMART, temperature, drive enclosure status (if it has a drive in it, if the door is locked and powered up) as well as various other inputs. SAF-TE supports SCSI, SATA and IDE based RAID array statuses, meaning it’s fully compatible with the controller and not just a patched job for status monitoring.
At the end of the card are the SATA ports stacked on top of each other in rows of two. This allows using the minimum amount of PCB space while fitting the entire amount of ports in and a small pezos buzzer is included to warn the user if a fault has occurred.
