# Delta DCIM — U-Level Granularity
Delta's DCIM (Data Center Infrastructure Management) platform monitors down to individual U positions in a rack, not just rack-level aggregates.
**What this means:**
Traditional monitoring (Zabbix, Nagios) polls infrastructure at the device level — you know a server is up, its CPU usage, maybe its inlet temperature. Delta's DCIM goes deeper: it knows which U slot each device occupies, its power draw, thermal contribution, and health status at that specific position.
**How it works:**
- IPMI (Intelligent Platform Management Interface) integration via ASRock AST2600 BMC
- Each server reports hardware telemetry: temperatures, fan speeds, voltages, power consumption
- The DCIM maps this telemetry to physical rack position (U1 through U42)
- Dashboard shows per-U power, thermal, and health data
**Why U-level matters:**
- **Capacity planning:** Know exactly which U slots are available and what their power/thermal budget is before placing a new server
- **Thermal management:** Identify hot spots at specific rack positions, not just "rack 7 is warm"
- **Troubleshooting:** When something overheats, you know it's U23, not "somewhere in the middle of the rack"
- **Asset tracking:** Physical location of every device is always current, automatically
**What it replaces:** Zabbix-style polling with manual asset spreadsheets. The DCIM combines infrastructure monitoring and physical asset management into one system built specifically for data center operations.
---
See also: [[Modular Data Centers MoC]]