Quanta AI Server Configuration Parameters

Quanta AI servers can be configured at both hardware and software levels, including BIOS, chipset, server management, and AI-specific transport protocols for optimal performance.

Hardware Configuration

QuantaGrid S31A-1U: This 1U server supports Intel® Xeon® E3-1200 v5 processors and the Intel® C236 chipset. Key hardware parameters include:

• Storage: 4x 3.5” HDD bays, 2 mini-SAS HD connectors supporting 8x SATA 6Gb/s ports, and 2x M.2 connectors for SATA or PCIe SSDs. Optional storage controllers are available for advanced setups.
• Power Supply: Available in fixed or redundant PSU models.
• Server Management: BMC interface allows configuration of system agent (SA), PCH-IO, and FRB-2 timers. BIOS screens provide boot order, iSCSI parameters, and onboard device configuration .

QuantaGrid D74A-7U: Designed for extreme AI and HPC workloads, this 7U server supports:

• GPUs: 8x NVIDIA H100/H200 or 8x AMD MI300X GPUs.
• Storage: Fast NVMe drives optimized for GPUDirect.
• Networking: Multiple GPUDirect RDMA-compatible NICs.
• Modularity: Tool-less chassis and tray design for easy service and future upgrades .

Software and AI Transport Configuration

Quanta AI servers use Agent-to-Agent (A2A) server configuration patterns for AI workloads, supporting multiple transport protocols:

• HTTP: Use uvicorn.run("main:app", host="0.0.0.0", port=8000, reload=True) with CORS middleware configured for allowed origins, credentials, methods, and headers .
• STDIO: Standard input/output transport for lightweight agent communication.
• SSE (Server-Sent Events): Implemented with EventSourceResponse(event_generator()) for streaming events.
• WebSocket: Supports real-time bidirectional communication with await websocket.accept() .

Configuration Templates: Quanta AI provides Python and TypeScript templates for MCP (Model Context Protocol) server setup. Placeholders follow the format {service}_{env}_your_key_here and should be replaced with environment-specific keys. Configuration should be applied according to the transport type and workload requirements .

BIOS and Server Management Parameters

• System Agent (SA) Configuration: Adjust CPU, memory, and chipset-related parameters.
• PCH-IO Configuration: Configure peripheral and I/O options.
• FRB-2 Timer: Set between 3–6 minutes for system recovery.
• Boot Options: Define boot order for devices; if no bootable device is detected, the system defaults to BIOS setup .

Best Practices

• Ensure rack stability and follow UL safety guidelines when installing servers. Anchor racks to prevent tipping.
• Use mechanical assistance for moving heavy servers.
• Apply CORS and security settings carefully for HTTP/SSE/WebSocket servers to prevent unauthorized access.
• For AI workloads, match GPU, storage, and network configurations to the expected model size and data throughput requirements .

This combination of hardware, BIOS, server management, and AI transport configuration ensures Quanta AI servers are optimized for high-performance AI and HPC workloads.