Intel X5460 AT80574KJ087N Data Sheet
Product codes
AT80574KJ087N
Thermal Specifications
96
6.3.2
PECI Specifications
6.3.2.1
PECI Device Address
The PECI device address for socket 0 is 0x30 and socket 1 is 0x31. Please note that
each address also supports two domains (Domain0 and Domain1). For more
information on PECI domains, please refer to the Platform Environment Control
Interface (PECI) Specification.
each address also supports two domains (Domain0 and Domain1). For more
information on PECI domains, please refer to the Platform Environment Control
Interface (PECI) Specification.
6.3.2.2
PECI Command Support
PECI command support is covered in detail in Platform Environment Control Interface
Specification. Please refer to this document for details on supported PECI command
function and codes.
Specification. Please refer to this document for details on supported PECI command
function and codes.
6.3.2.3
PECI Fault Handling Requirements
PECI is largely a fault tolerant interface, including noise immunity and error checking
improvements over other comparable industry standard interfaces. The PECI client is
as reliable as the device that it is embedded in, and thus given operating conditions
that fall under the specification, the PECI will always respond to requests and the
protocol itself can be relied upon to detect any transmission failures. There are,
however, certain scenarios where PECI is known to be unresponsive.
improvements over other comparable industry standard interfaces. The PECI client is
as reliable as the device that it is embedded in, and thus given operating conditions
that fall under the specification, the PECI will always respond to requests and the
protocol itself can be relied upon to detect any transmission failures. There are,
however, certain scenarios where PECI is known to be unresponsive.
Prior to a power on RESET# and during RESET# assertion, PECI is not guaranteed to
provide reliable thermal data. System designs should implement a default power-on
condition that ensures proper processor operation during the time frame when reliable
data is not available via PECI.
provide reliable thermal data. System designs should implement a default power-on
condition that ensures proper processor operation during the time frame when reliable
data is not available via PECI.
To protect platforms from potential operational or safety issues due to an abnormal
condition on PECI, the Host controller should take action to protect the system from
possible damage. It is recommended that the PECI host controller take appropriate
action to protect the client processor device if valid temperature readings have not
been obtained in response to three consecutive gettemp()s or for a one second time
interval. The host controller may also implement an alert to software in the event of a
critical or continuous fault condition.
condition on PECI, the Host controller should take action to protect the system from
possible damage. It is recommended that the PECI host controller take appropriate
action to protect the client processor device if valid temperature readings have not
been obtained in response to three consecutive gettemp()s or for a one second time
interval. The host controller may also implement an alert to software in the event of a
critical or continuous fault condition.
6.3.2.4
PECI GetTemp0() and GetTemp1() Error Code Support
The error codes supported for the processor GetTemp0() and GetTemp1() commands
are listed in
are listed in
below:
§
Table 6-12. GetTemp0() GetTemp1()Error Codes
Error Code
Description
0x8000h
General sensor error
0x8002h
Sensor is operational, but has detected a temperature below its operational range
(underflow).