Intel
®
81341 and 81342—DDR SDRAM Memory Controller
Intel
®
81341 and 81342 I/O Processors
Developer’s Manual
December 2007
620
Order Number: 315037-002US
illustrates the DDR SDRAM waveforms after the assertion of
P_RST#
during
a true power failure.
CKE[1:0]
must be held low throughout the power-down period. The memory controller
drives it low initially with the
self-refresh
command, but an external pull-down is
required to continually drive it low when the 81341 and 81342 loses power. External
logic ensures that
CKE[1:0]
is held low after the memory controller initially deasserts
it. Likewise, the external logic must stop driving
CKE[1:0]
low once
P_RST#
is
deasserted by the system.
Due to the high loading on
CKE
and the requirement of PC200 DDR operation, the
memory controller must drive two copies to the DDR SDRAM DIMM. The board layout
distributes the two
CKE[1:0]
signals between the two DDR SDRAM banks equally.
As long as the DDR SDRAM memory subsystem is powered with a battery source and
CKE[1:0]
is held low, the DDR SDRAM preserves its memory image.
When power is restored, the system asserts
P_RST#
to the 81341 and 81342. While
the 81341 and 81342 is reset,
CKE[1:0]
is held low by memory controller. After
P_RST#
is deasserted (and subsequently, internal bus reset is deasserted), the 81341
and 81342 must be re-initialized to reset the DDR SDRAM memory subsystem
operating parameters. The first step of DDR SDRAM initialization sequence re-asserts
CKE[1:0]
to ones and the memory controller resumes refreshing. DDR SDRAM
initialization sequence does not affect memory contents. For more details about the
DDR SDRAM initialization sequence, refer to
Note:
The power failure mechanism in the memory controller is not responsible for
maintaining the 81341 and 81342 state. The purpose of this mechanism is to maintain
the memory so that any data cached in the local memory can be flushed once power is
restored. Any data queued within the 81341 and 81342 components (DMCU, ATU,
DMAs, etc.) is lost.
Figure 98. Power Failure Sequence
RAS#
CAS#
WE#
MA[10]
0
1
2
3
4
5
6
7
8
9
10 11 12
M_CK
13 14 15
Trp
16 17 18 19
Reset
CE[0]#
due to Power Fail
Wait for 13 to 15 clocks
P_RST#
Precharge-All Self-Refresh
CKE[0]
CE[1]#
20
CKE[1]
B6265-01