Scylla AI - AST-48-9148 - Scylla AI Inference Server, 2U, 1TB

Product Overview

The Exacq AST-48-9148 is a high-performance 2U Scylla Vanguard AI inference server for demanding large-scale video analytics deployments. Its 48-core CPU, 91 TFLOPS AI inference performance, and 48GB VRAM deliver maximum analytics throughput in a space-efficient rackmount chassis.

Key Configuration

Model: AST-48-9148 | Form Factor: 2U rackmount | CPU: 48-core | AI Performance: 91 TFLOPS | VRAM: 48GB | Storage: 1TB | Software: Scylla AI inference platform

Key Features

• · 2U rackmount AI inference server
• · 48-core CPU for high parallel processing capacity
• · 91 TFLOPS AI inference performance
• · 48GB GPU VRAM for concurrent analytics inference
• · 1TB onboard storage
• · Highest AI inference performance tier in the 2U AST lineup
• · Supports simultaneous processing across Scylla AI analytics modules
• · Combines high CPU parallelism with GPU inference capacity
• · Space-efficient 2U chassis for data-center deployments

Technical Context

The AST-48-9148 performs Scylla AI inference locally while its 48-core CPU handles high-channel-count video processing and analytics preprocessing. Its 91 TFLOPS GPU tier maximizes concurrent inference capacity without requiring a larger 4U chassis.

Installation Notes

Install the server in a suitable ventilated 2U rack position with network connectivity to cameras and the exacqVision VMS environment. Configure Scylla licenses, camera assignments, storage, and analytics rules before operation.