QUASAR Ryzen AI Max Edge Inference Node

Overview

A compact AI node that still has a slot. Most small-form AI boxes are sealed — what ships is what you get. This one carries a PCIe ×16 expansion slot (4 lanes of PCIe 4.0) on a daughterboard, so a capture card, an accelerator or additional I/O can go in at the edge. Behind it sits the Ryzen AI Max platform with up to 128 GB of unified memory.

Key highlights

• PCIe ×16 slot (4× PCIe 4.0) — capture card, NPU card or additional I/O at the edge.
• Up to 128 GB unified memory, up to 96 GB assignable to graphics.
• 50 TOPS NPU plus a Radeon 8060S iGPU for on-site inference.
• Alternate daughterboards: 6× SATA for local capture retention, or additional NVMe.
• Rear-I/O options including Oculink, optical, USB, HDMI and DP.
• x86 and Windows — the same management and imaging tools as the rest of the estate.
• Compact enough for a cabinet, a control room or a plant-floor enclosure.

AI workload fit

• On-site inference where data cannot leave the location.
• Computer vision against local cameras via a capture card.
• RAG over site-local documents and records.
• Agentic AI running unattended at a remote location.
• NLP & speech processing of local audio.
• Branch, plant, clinic and campus deployments needing autonomous AI.

AI workload positioning

Edge deployments fail on I/O and serviceability far more often than on compute. A sealed mini cannot take a capture card, cannot add retention storage, and cannot be adapted when the site requirement changes six months in. The expansion slot is the entire point of this configuration — the AI platform is the same one in the desk-side units. Specify the daughterboard with the deployment, not after it.

Industry use cases

• Manufacturing: plant-floor vision inspection with local camera capture.
• Public sector & sovereign: site-local AI at offices and facilities, on soil.
• Healthcare: departmental inference kept inside the hospital network.
• BFSI: branch-level processing where data must not traverse the WAN.
• Enterprise & GCCs: remote-site AI with central imaging and management.

Performance & how to be sure

Edge performance is decided by the I/O path as much as the accelerator — camera ingest, local retention and network return. RDP sizes those together rather than quoting a compute figure in isolation. Rather than quote a tokens/sec figure that will not match your workload, RDP offers a “benchmark your model” session: bring the model, precision and context length you actually intend to run, and we will validate it on this exact configuration before you commit.

Series & upgrade path

QUASAR tier in Edge & Micro Data Center. Below: RDP CARINA edge AI nodes for lighter single-purpose duty. Beside: the desk-side CARINA/QUASAR Ryzen AI Max units where expansion is not needed. Above: RDP micro data center pods when a site needs rack-scale capacity.

On-prem vs cloud (TCO)

For sustained daily AI work this configuration removes per-GPU-hour billing, queueing for scarce instances, and egress charges on your own data — and keeps everything on-premises for DPDP and data-residency obligations. Cloud still wins for burst capacity and one-off very large training runs; on-prem wins on sustained utilisation, control and predictable INR capital cost. We model the crossover with your actual usage rather than assert it.

Software & day-one readiness

Ships with Windows 11 or Ubuntu and the AMD AI stack — ROCm, DirectML and ONNX Runtime — with local model runtimes (LM Studio, Ollama, llama.cpp) validated before dispatch. The GPU draws from the same unified pool as the CPU, so you assign memory to the model rather than shopping for a card with enough VRAM. Migration from a CUDA workflow is a real exercise and RDP scopes it honestly rather than calling it a drop-in swap.

Power, thermal & acoustics

Compact chassis with internal or external supply depending on the daughterboard and slot population. Air-cooled. Enclosure, mounting, ingress protection and operating-temperature requirements are specified per site at quotation rather than assumed. Exact wattage, BTU and dB(A) figures come from RDP bench measurement rather than estimates — ask for the site-readiness sheet with your quote.

Deployment, warranty & support

Built to order by RDP Technologies. Supplied with GST invoice, pan-India onsite support, and availability through GeM for public-sector procurement. Built to order — typically 6–8 weeks; daughterboard and rear-I/O configuration confirmed at quotation.

Why RDP

RDP Technologies is a Make-in-India OEM with 14+ years and 300,000+ units shipped, supplying AI infrastructure from desk-side systems to rack-scale AI factories — with predictable INR pricing, GST invoicing and pan-India onsite service.

Buy with confidence

Use Request a Quote to reach an RDP solution architect for sizing, a benchmark-your-model session, financing options and a delivery plan. No obligation.

Specifications