Seagate Nytro 2.5” x 7mm SSD Product Manual, Rev. A
7
Scope
www.seagate.com
1.6
Out of Band Management (SMBus)
The provides sub-system management data to the host over the SMBus interface at 0x6A
.
Table 5 Out of Band Management (SMBus) Protocol
Command
Code
Offset
(byte)
Description
0
00
Length of Status:
Indicates number of additional bytes to read before encountering PEC. This value should always
be 6 (06h) in implementations of this version of the spec.
01
Status Flags (SFLGS):
This field indicates the status of the NVM subsystem.
SMBus Arbitration
– Bit 7 is set ‘1’ after a SMBus block read is completed all the way to the stop bit without bus
contention and cleared to ‘0’ if a SMBus Send Byte FFh is received on this SMBus slave address.
Drive Not Ready
– Bit 6 is set to ‘1’ when the subsystem cannot process NVMe management commands, and the rest of
the transmission may be invalid. If cleared to ‘0’, then the NVM subsystem is fully powered and ready to respond to man-
agement commands. This logic level intentionally identifies and prioritizes powered up and ready drives over their pow-
ered off neighbors on the same SMBus segment.
Drive Functional
– Bit 5 is set to ‘1’ to indicate an NVM subsystem is functional. If cleared to ‘0’, then there is an
unrecoverable failure in the NVM subsystem and the rest of the transmission may be invalid.
Reset Not Required
- Bit 4 is set to ‘1’ to indicate the NVM subsystem does not need a reset to resume normal
operation. If cleared to ‘0’ then the NVM subsystem has experienced an error that prevents continued normal
operation. A controller reset is required to resume normal operation.
Port 0 PCIe Link Active
- Bit 3 is set to ‘1’ to indicate the first port’s PCIe link is up (i.e., the Data Link Control and
Management State Machine is in the DL_Active state). If cleared to ‘0’, then the PCIe link is down.
Port 1 PCIe Link Active
- Bit 2 is set to ‘1’ to indicate the second port’s PCIe link is up. If cleared to ‘0’, then the second
port’s PCIe link is down or not present.
Bits 1-0 shall be set to ‘1’.
02
SMART Warnings:
This field shall contain the Critical Warning field (byte 0) of the NVMe SMART / Health Information
log. Each bit in this field shall be inverted from the NVMe definition (i.e., the management interface shall indicate a
‘0’ value while the corresponding bit is ‘1’ in the log page). See the NVMe specification for bit definitions.
If there are multiple controllers in the NVM subsystem, the management endpoint shall combine the Critical Warning
field from every controller such that a bit in this field is:
Cleared to ‘0’ if any controller in the subsystem indicates a critical warning for that corresponding bit.
Set to ‘1’ if all controllers in the NVM subsystem do not indicate a critical warning for the corresponding bit.
03
Composite Temperature (CTemp):
This field indicates the current temperature in degrees Celsius. If a temperature
value is reported, it should be the same temperature as the Composite Temperature from the SMART log of hottest con-
troller in the NVM subsystem. The reported temperature range is vendor specific, and shall not exceed the
range -60 to +127°C. The 8 bit format of the data is shown below.
This field should not report a temperature when that is older than 5 seconds. If recent data is not available, the
NVMe management endpoint should indicate a value of 80h for this field.
Value Description
00h-7Eh Temperature is measured in degrees Celsius (0 to 126C)
7Fh 127C or higher
80h No temperature data or temperature data is more the 5 seconds old.
81h Temperature sensor failure
82h-C3h Reserved
C4 Temperature is -60C or lower
C5-FFh Temperature measured in degrees Celsius is represented in twos complement (-1 to -59C)
04
Percentage Drive Life Used (PDLU):
Contains a vendor specific estimate of the percentage of NVM subsystem NVM life
used based on the actual usage and the manufacturer’s prediction of NVM life. If an NVM subsystem has multiple con-
trollers the highest value is returned. A value of 100 indicates that the estimated endurance of the NVM in the NVM sub-
system has been consumed, but may not indicate an NVM subsystem failure. The value is allowed to exceed 100.
Percentages greater than 254 shall be represented as 255. This value should be updated once per power-on hour and
equal the Percentage Used value in the NVMe SMART Health Log Page.
06 : 05
Reserved:
Shall be set to 0000h.
07
PEC:
An 8 bit CRC calculated over the slave address, command code, second slave address and returned data.
Algorithm is in SMBus Specifications.
