CARINA Ryzen AI Max 2L Compact AI PC

Overview

The same silicon, more thermal headroom. A two-litre chassis lets the Ryzen AI Max+ 395 run at a 120–132 W envelope rather than throttling back, which is what turns a demo machine into one that holds throughput through a working day. Up to 128 GB unified memory, full desktop I/O, and a footprint that still disappears behind a monitor.

Key highlights

• 120–132 W sustained envelope — headroom the 1 L class does not have.
• Up to 128 GB LPDDR5X-8000, 256-bit — up to 96 GB assignable to graphics.
• AMD Ryzen AI Max+ 395, 16C/32T, with a 40-CU Radeon 8060S and 50 TOPS NPU.
• Full desktop I/O — USB4, 3× USB-A, HDMI 2.1 FRL, DP 1.4, 2.5GbE, SD card.
• Choice of internal PSU or 240 W external adapter depending on deployment.
• 2 L, 203 × 192 × 70 mm — VESA-mountable behind a display.
• Windows 11 or Ubuntu, imaged and validated by RDP before dispatch.

AI workload fit

• Sustained local inference through a full working day.
• RAG against private corpora held on the machine.
• Agentic AI workflows running continuously rather than on demand.
• Generative AI for document, image and text production.
• NLP & speech — transcription, summarisation, meeting intelligence.
• Light fine-tuning and adapter training on small models.

AI workload positioning

Between the 1 L unit and the liquid-cooled 2 L, this is the volume choice: enough thermal budget to hold clocks under a real workload, in a chassis that still belongs on a desk rather than in a rack. As with every machine in this family, the honest framing is inference and development, not training — the unified pool decides what fits, and the NPU is not a substitute for a discrete accelerator when you need sustained training throughput.

Industry use cases

• Enterprise & GCCs: standardised per-desk AI hardware across an applied-AI team.
• BFSI: analyst desktops running local models over regulated data.
• Healthcare: on-premises clinical documentation and coding assistance.
• Public sector & sovereign: departmental AI with no external dependency.
• Research & education: lab and teaching machines that run real models.

Performance & how to be sure

The extra litre buys sustained rather than peak performance — the difference shows in the second hour of a workload, not the first minute. 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

CARINA entry tier. Below: the 1L Mini AI Workstation. Above: the QUASAR 2L Liquid-Cooled AI PC at 147 W, then the 4L Desktop Workstation, then RDP AI Workstations with discrete RTX PRO GPUs for genuine fine-tuning throughput.

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. Because the GPU draws from the same unified pool as the CPU, 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

Internal PSU or a 240 W external adapter on a standard socket. Air-cooled with a desk-appropriate acoustic profile; no facility requirements. 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 4–6 weeks.

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