Park Place adds GPU and liquid cooling monitoring to Entuity

Park Place Technologies has updated Entuity Software to give enterprise IT teams a consolidated view of AI accelerators, interconnect fabrics and cooling equipment. Entuity V23.0 P03 supports GPUs from AMD, Intel and NVIDIA, combining hardware health and performance data to help teams identify problems before they cause outages.

The release extends visibility beyond operating-system-level monitoring, tracking accelerator temperature, power draw, load and health status. It also monitors direct-to-chip liquid cooling, including pump flow, coolant temperature, pressure, reservoir levels, valve positions and leak detection. Correlating cooling conditions with GPU telemetry can highlight risks of throttling, instability or downtime.

Fault detection focuses on grey failures, which can gradually reduce performance without triggering conventional availability alerts. Memory health indicators, ECC errors and switch-level interconnect monitoring help expose hardware degradation and deteriorating links. Visibility into GPU partitions, accelerator instances, memory usage and utilisation also helps teams identify underused resources. The release includes 14 thresholds enabled by default and about 415 telemetry streams across accelerators and cooling.

Cooling support covers suppliers including Accelsius, GRC, nVent and ZutaCore. In-band and out-of-band Redfish support through iDRAC and iLO prioritises in-band monitoring to avoid duplicate GPU counts. Existing Entuity customers with active maintenance contracts can upgrade at no additional cost, while new users can deploy directly onto the current release.