I have read carefuly the technical spec. of the H67 chip, and found nowhere port multiplier reference.
The same within the motherboard specs (DH67GD), where it is written that the sata ports "support only one peripheral", we know that, maybe it is an elegant manner to write that they do not support port multiplier.
And, finaly, if you read the help of Intel Rapid Storage console, it is clearly written that it does not support port muliplier.
I dont know if the limitation is only software.
Now, with USB3, I am not confident that e-sata port multiplier have a great future.
Silicon Image chip on PCIe may be a workaround.
That is a bit strange.
What model "5 bay SATA drive" are you using?
Drive expanstion bays are normally either passive requiring a connection for each HDD or active which have SAS\SATA expanders on the expander backplane which act as small processors to negotiate the and manage the down stream devices.
Since the first drive is showing up, it indicates the motherboard is incomunications with the backplane / expander to the drive.
The remaining drives not showing up, would then (most likley) be a connunication failure on between the expander and the drives or you have a passive backplane that would have 5 connectors on it to connect to the mother board.
This behaviour is typical of multi drive e-sata bays when used in JBOD mode (several independant units, single port connection), needing an e-sata port with the port multiplier feature. (included in sata specification, but not mandatory).
When the port is not capable, only one drive is visible.
Often these bays may also run in RAID modes, or what some call BIG mode, where several disks are combined to make a single volume, in this case port muliplier feature is not required.
Inside the bay there is a kind of RAID controller, or just switch, with one host port and several sata ports for the drives.
Silicon Image is making such controllers, and also provides PCIe cards with a port multiplier capable e-sata interface. (SIL 3112 chip, maybe others).
I'm using a Towerraid box running in JBOD mode with eSATA and USB 2.0. All of the drives appear when I plug in via USB 2.0 and when I plug in on another computer with eSATA which I know has port multiplier. So the box itself is working fine, the issue is clear on the Intel chipset side.
I'm somewhat surprised that the Intel chipset does not support port multiplier natively -- one would think that Intel would implement the entire eSATA standard instead of leaving parts off. Of course USB3.0 is an option but I happen to like the enclosure that I'm using and the same vendor's USB3.0 enclosure is annoyingly noisy so I was hoping not to use it.
I'm going to install a PCI eSATA card but there's only one PCI slot so that does limit potential for future expansion.
Intel, if there is a way of fixing this via a BIOS or software update, I think it would be very helpful.
I was not aware of the "port multiplier" feature serving as a software multiplexing feature spitting the band width between several drives.
I have always ran in to hardware expanders.
SATA debuging just got one more level of completity.
I have a similar problem with an Asrock P67 Extreme 6 board. I used to connect my Fantec QB-35US3R usb3/esata enclosure on a P35 system through an esata cable on an esatabracket connected to a sataport. No usb3 on that board yet. Passthru of about 240 MB/s. My Fantec does not get recognized on the same bracket connected to a sataport on my P67 board. In bios (uefi) the Fantec is registered, but in windows the Fantec is gone. I'm running it in raid 0. Because port multiplier is not supported, none of the four drives in the Fantec gets noticed. Now I have to use the esata port from the Marvell 9120 chipset. This (sadly lacking) chipset does support port multiplier, and the Fantec is recognized instantly. Downside is the slow passthru of about 140 MB/s. Usb3 is a dead end. It seems the usb to sata bridge in the Fantec (Jmicron JMS539) has issues with the usbcontroller. The JMS539 has issues with all usb3 controllers. The drive gets connected and disconnected, or does not get recognized at all. Or works for a while at usb2 speeds. All in all the P35 southbridge seems to be superior to the one on the P67 board.
Currently, port multipliers are supported on the chipset side, on Intel® ICH10R, controllers. However, the functionality has not been fully implemented to the Intel(R) Rapid Storage Technology (driver) yet. We expect a new version of the software will add full support for the feature.
It's been about a year now since you posted that port multiplier support will be added. Any update?
P.S. If Intel had a fraction of the mojo for software that it does for hardware, we wouldn't be talking about things like this on boards.
Hello, Jose! What's new? Any update? Thank you!
Intel® Rapid Storage does not support port multiplier and there are no plans to include such feature at this moment. I am sorry for the inconvenience.
Not supporting this feature doesn't strike me as a really good way of highlighting the bleeding edge technical capabilities of Intel. How come a $5 controller from China can do it, but the Intel one in my $2,000 laptop can't? This is an excellent example of why Intel is slipping. Keep it up, and we'll all be using ARM based stuff because there was no more reason to pay $2,000 for a device that can't do something so obvious.
To say we are clamouring for SATA PMP support in the Windows IRST driver is an understatement.
Your idea is interesting, I will transfer your request to my manager; hopefully we will have the option in the future.