ASUS ESC4000A-E10 SEL Reference

IPMI System Event Log reference for ASUS ESC4000A-E10 GPU servers with AMD EPYC processors.

Platform: ASUS ESC4000A-E10

BMC: ASMB10-iKVM

Version: 2.0x

Overview

The ASUS ESC4000A-E10 is a 4U GPU server supporting up to 8 AMD EPYC processors and multiple NVIDIA GPUs. The ASMB10-iKVM BMC provides IPMI 2.0 management capabilities.

Platform-Specific Sensors

Power Bus Monitoring

The ESC4000A-E10 has dedicated power bus sensors (PMBPower) that monitor the internal power distribution.

Sensor	Description	Function
PMBPower1	Power Module Bus 1	Primary power rail monitoring
PMBPower2	Power Module Bus 2	Secondary power rail monitoring

PMBPower Event Types

Event	Severity	Meaning	Action
Upper Critical going high	🔴 Critical	Power draw exceeds safe limit	Check GPU workload, verify PSU capacity
Upper Non-critical going high	🟡 Warning	Power approaching limit	Monitor workload
Lower Critical going low	🔴 Critical	Power drop detected	Check PSU connections
Lower Non-critical going low	🟡 Warning	Power slightly low	Verify PSU health
Deasserted	🟢 Info	Condition cleared	Normal operation

Note: PMBPower events often correlate with high GPU utilization. Frequent events may indicate insufficient PSU capacity for the workload.

Thermal Zones (TR1, TR3)

The ESC4000A-E10 uses thermal zone sensors to monitor component temperatures.

Sensor	Location	Normal Range
TR1 Temperature	CPU/VRM thermal zone	< 85°C
TR3 Temperature	GPU/PCIe thermal zone	< 90°C

Temperature Thresholds

Threshold	TR1 (CPU Zone)	TR3 (GPU Zone)
Upper Non-Critical	75°C	80°C
Upper Critical	85°C	90°C
Upper Non-Recoverable	95°C	100°C

Troubleshooting High TR Temperatures

Check fan operation - Verify all chassis and GPU fans are spinning
Verify airflow - Ensure proper cable management, no obstructions
Check ambient temperature - Datacenter HVAC within spec?
Review workload - GPU-intensive tasks will raise TR3
Clean heatsinks - Dust accumulation reduces cooling efficiency

AMD EPYC Specific Voltages

These sensors monitor AMD EPYC-specific power rails.

Sensor	Description	Normal Range
+VCORE1	CPU Core voltage	Per processor spec (0.9V - 1.2V typical)
+VSOC1	SoC (System-on-Chip) voltage	~0.9V

Voltage Event Interpretation

Event	Cause	Action
Upper Critical going high	Over-voltage condition	Check VRM, may indicate failure
Upper Non-critical going high	Voltage slightly high	Monitor, usually transient
Lower Critical going low	Under-voltage condition	Check PSU, power cables
Lower Non-critical going low	Voltage slightly low	Monitor, check load

Warning: Voltage deviations outside ±5% can cause system instability or component damage.

Drive Bay Sensors

The ESC4000A-E10 has a hot-swap drive backplane with individual bay sensors.

Sensor	Location
Backplane1 HD05	Drive bay 5
Backplane1 HD07	Drive bay 7
Backplane1 HD08	Drive bay 8

Drive Bay Events

Event	Meaning	Action
Drive Present \| Asserted	Drive inserted	Normal hot-swap
Drive Present \| Deasserted	Drive removed	Verify intentional

Memory Events

Memory Training Errors

Sensor	Description	Severity
Memory_Train_ERR	Memory training error during POST	🟡 Warning

Memory training errors occur during system boot when the memory controller calibrates timing for the DIMMs.

Common Causes:

DIMM not fully seated
Incompatible DIMM
Failed DIMM
Memory controller issue

Troubleshooting:

Reseat the affected DIMM
Run memory diagnostics (memtest86+)
Check DIMM compatibility with server QVL
Try DIMM in different slot
Replace DIMM if errors persist

ECC Error DIMM Identification

The ESC4000A-E10 reports ECC errors with DIMM location:

Memory | Correctable ECC | Asserted | **DIMM A1 (CPU1)** | [CPU1_ECC1]

Field	Meaning
DIMM A1	Physical DIMM slot
(CPU1)	Associated CPU socket
[CPU1_ECC1]	ECC controller identifier

DIMM Slot Layout (per CPU):

Slots A1-A8: CPU1 memory channels
Slots B1-B8: CPU2 memory channels (if dual-socket)

Power Supply Sensors

The ESC4000A-E10 supports redundant power supplies with individual monitoring.

Sensor	Description
PSU1 AC Lost	PSU 1 AC power status
PSU2 AC Lost	PSU 2 AC power status
PSU2 PWR Detect	PSU 2 presence and power detection
REDUNDANCY_PSU	Redundancy status
PSU1 Over Temp	PSU 1 thermal status
PSU2 Over Temp	PSU 2 thermal status

PSU Events and Actions

Event	Severity	Cause	Action
AC lost \| Asserted	🔴 Critical	Power cord unplugged or PDU failure	Check PDU, power cord
AC lost \| Deasserted	🟢 Info	Power restored	Normal operation
Failure detected \| Asserted	🔴 Critical	PSU hardware failure	Replace PSU
Fully Redundant \| Asserted	🟢 Info	Both PSUs healthy	Normal
Redundancy Lost \| Asserted	🟡 Warning	Only one PSU online	Check/replace failed PSU
Over Temp \| Asserted	🔴 Critical	PSU overheating	Check airflow, replace if persistent

Standard Voltage Rails

Sensor	Normal	Warning	Critical
+3.3V	3.135V - 3.465V	±5%	±10%
+3.3VSB	3.135V - 3.465V	±5%	±10%
+5VSB	4.75V - 5.25V	±5%	±10%

Critical Interrupt Events

Event	Severity	Cause	Action
Bus Correctable error	🔴 Critical	PCIe/memory bus error (corrected)	Monitor frequency
Bus Fatal Error	🔴 Critical	Unrecoverable bus error	Check PCIe cards, reseat GPUs
PCI SERR	🟢 Info	PCIe system error	Usually transient

Recommended Maintenance

Monthly

Clear SEL if approaching limit
Review ECC error trends
Check thermal zone temperatures

Quarterly

Clean air filters and heatsinks
Verify fan operation
Check PSU status

Annually

Firmware updates (BMC, BIOS)
Full system diagnostics
Thermal paste replacement if needed

Last updated: December 2025