Pci gen2 does not work

[amtony....@gmail.com]

I used S7010, it's don't have sata3 port

Then I just brought a pci to sata3 card.

And I use pcie 8x , They are gen2 pcie port but when i boot, the System say used pci gen 1 and speed it's gen1

This card can used with another m/b and its say used pci gen2 , how should i do?

[Paul]

The manual mentions Tylersburg 36D/24D, but I assume it's
actually using 36D, in order to match the S7010 slot combinations.

(S7010 user manual)
http://www.tyan.com/manuals/S7010_UG_v1.1a_06212012.pdf

Tylersburg 36D (5520)

x16 Gen2
x8 Gen2
x8 Gen2
x4 Gen2 --- routed to Marvel 6440 as x4 Gen1 <--- anomaly

ICH10 (up to 6 Gen1)
x4 Gen1
x1 Gen1 82574 NIC
x1 Gen1 82574 NIC

The Tylersburg has flexible lane configurations. More
flexible than some desktops in terms of being able to
chop up the interface into smaller lane groups. The
diagram here shows some of the lane group options.

http://techreport.com/review/16656/intel-xeon-w5580-processors/2

Now, the weird part, is Tyan makes the P3301,
and the specification for Marvell 88SE6440 says...

http://www.tyan.com/HBA_SKU_spec.aspx?ProductType=HBA&pid=7&SKU=600000148

PCI-E Gen2 x4 host interface

So that's the first anomaly I see. Why is that
x4 group claimed to be running at Gen 1 speed
to the 6440 ?

One of the requirements for Gen2 slots, is a
"low jitter" clock. Sometimes, when a motherboard
designer attempts to use some Intel chipsets,
there are insufficient low jitter Gen2 clock
signals, to make it possible to run a slot
in Gen2 mode. Instead, a regular Gen1-quality
clock is routed to the slot, and (presumably by
BIOS), the slot is declared Gen 1.

The TechReport article above, the picture shows
me that the Tylersburg supports operation as
groups of x4 Gen2 slots on the main PCI-E
connectors. So there should have been plenty
of the correct kind of clock signal, on the
chipset itself.

The lanes coming from the Southbridge are only
Gen1, so the claims there are perfectly consistent
with normal experience. Only Gen 1 comes off
the Southbridge (and that's also limited
typically by the DMI bus).

And there is nothing in the BIOS release notes
to suggest they fixed any PCI Express issues that
way.

http://www.tyan.com/support_download_bios.aspx?model=S.S7010

You would have to check with Tyan support, to see
if this behavior is "normal". Seeing the 6440
running in Gen 1, according to the manual,
is the only hint I see that something is
abnormal about the design. I'm sure the technical
writers check these documents very carefully.

Paul