Scylla AI - AST-24-2620 - AI Inference Server, 4U, 1TB

Product Overview

The Exacq AST-24-2620 is a 4U Scylla Vanguard AI inference server designed for demanding, expandable video analytics deployments. Its 24-core CPU and 20 TFLOPS AI inference performance provide substantial processing capacity while the larger chassis supports storage expansion, redundancy, and enhanced thermal headroom.

Key Configuration

Model: AST-24-2620 | Form Factor: 4U rackmount | CPU: 24-core | AI Performance: 20 TFLOPS | Storage: 1TB | Software: Scylla AI pre-configured for exacqVision VMS integration

Key Features

• · 4U rackmount AI inference server
• · 24-core processing architecture
• · 20 TFLOPS AI inference performance
• · 1TB onboard storage
• · Pre-configured for Scylla AI analytics software
• · Native integration with exacqVision VMS
• · More CPU cores than the 16-core AST-16-2620 platform
• · Expanded chassis supports additional drive bays for storage expansion
• · Supports redundant power supply options and enhanced thermal headroom
• · Pairs with Scylla analytics licenses for camera deployments

Technical Context

The AST-24-2620 performs Scylla AI inference locally while providing additional CPU capacity for larger camera populations and demanding workloads. Its 4U enclosure offers more physical expansion space and thermal capacity than compact 1U systems at the same 20 TFLOPS inference tier.

Installation Notes

Install the server in a suitable ventilated 4U rack position with network connectivity to cameras and exacqVision VMS. Configure storage expansion, optional redundant power, Scylla licenses, camera assignments, and analytics rules before operation.