EMARQUE AI PRO 800 AI Server - RTX PRO 6000 Blackwell

The EMARQUE AI PRO 800 is EMARQUE's own-brand multi-GPU AI server, integrated and validated in Malaysia in a In Win IW-RG658-PRO 6.5U rackmount chassis (4.5U GPU chamber + 2U system chamber), with up to 8× NVIDIA RTX PRO 6000 Blackwell datacentre GPUs inside. Each unit ships as a complete system: the 6.5U chassis with redundant Titanium PSUs, 8× RTX PRO 6000 Blackwell GPUs (96 GB GDDR7 each), dual-socket server-class CPU, ECC system memory, enterprise NVMe storage, ConnectX-class networking, and Ubuntu LTS preinstalled with the full NVIDIA AI Enterprise stack ready to run.

For teams deploying serious on-prem AI, EMARQUE AI Solutions are available on request — deployment scoping, rack and PDU sizing, ConnectX-8 networking integration, multi-server clustering, Base Command and Kubernetes orchestration, private LLM / RAG / agent stack setup, team training. Final configuration (CPU SKU, system memory, NVMe capacity, networking SKU, software stack) confirmed on enquiry against your specific workload.

Indicative configuration — final configuration confirmed on enquiry:

• GPUs: up to 8× NVIDIA RTX PRO 6000 Blackwell Server Edition — 96 GB GDDR7 ECC per GPU (up to 768 GB total), 24,064 CUDA cores + 752 fifth-gen Tensor cores + 188 fourth-gen RT cores each (192,512 CUDA cores aggregate).
• AI compute: 4 PFLOPS FP4 per GPU — 32 PFLOPS FP4 aggregate; 120 TFLOPS FP32, 1.6 TB/s memory bandwidth per GPU.
• GPU class: PCIe Gen5 (no NVLink baseboard), 600W passive, dual-slot full-height full-length; Multi-Instance GPU (MIG) up to 4 instances per GPU. Optimised for LLM fine-tuning, inference and RAG (per-GPU model classes under ~70B parameters), plus vision, video and HPC.
• Chassis: In Win IW-RG658-PRO 6.5U rackmount (4.5U GPU + 2U system chamber), 650 × 490 × 289 mm, 19-inch standard rack.
• GPU support: up to 8 dual-slot full-height GPUs, 360 mm max length; 16 full-height PCIe Gen5 slots via MCIO converter board.
• CPU: dual-socket server-class (Intel Xeon or AMD EPYC; selectable per workload).
• System memory: ECC DDR5, scoped against workload.
• Storage: up to 8× E1.S hot-swap NVMe + 4× 2.5-inch bays.
• Networking: NVIDIA ConnectX-class SuperNIC up to 400 / 800 Gb/s; management 10 GbE.
• Power: 9,600W (3+1 redundant CRPS 3200W, 80 PLUS Titanium).
• Cooling: 14 PWM fans (10 GPU + 4 system), 17,000 RPM, hot-swap cages, front-to-back airflow, datacentre conditions.
• OS / Stack: Ubuntu LTS + full NVIDIA AI Enterprise stack.

Email business@emarque.co for the full configuration sheet.

The NVIDIA RTX PRO 6000 Blackwell Server Edition is NVIDIA's flagship Blackwell-generation datacentre GPU: 96 GB GDDR7 ECC, 24,064 CUDA cores, 752 fifth-gen Tensor cores with FP4 precision, 4 PFLOPS of FP4 AI performance and 1.6 TB/s of memory bandwidth, on a 600W passively-cooled PCIe Gen5 card. It supports Multi-Instance GPU (MIG) up to 4 isolated instances. A 8× configuration in the In Win IW-RG658-PRO 6.5U rackmount chassis gives 768 GB aggregate GPU memory and 32 PFLOPS of aggregate FP4 compute — the throughput to serve multi-tenant inference, fine-tune models (per-GPU classes under ~70B parameters), run organisation-scale RAG, and handle vision plus multimodal workloads, all from one box in a standard datacentre rack.

Because these are PCIe GPUs (not an NVLink baseboard like HGX B300), the tier is the most cost-effective and flexible path into multi-GPU on-prem AI — scale the GPU count to your workload. The 6.5U In Win chassis (4.5U GPU chamber + 2U system chamber) is purpose-built for dense GPU servers — dual-chamber thermal separation, hot-swap fans, redundant Titanium power, full 360 mm GPU length at full PCIe Gen5 bandwidth.

The EMARQUE AI PRO 800 ships with EMARQUE's 3-year warranty on the EMARQUE-built system (assembly, configuration, thermals), with individual component warranties (GPUs, PSU, motherboard, NVMe) per manufacturer on top — up to 3–5 years depending on the component. EMARQUE's built-in role is Malaysia-based local handling:

• Single point of contact — pre-sale, post-sale and warranty routed through the same EMARQUE team. Real engineer, no offshore queue.
• RMA routing — if a component fault appears, EMARQUE handles the warranty claim path on your behalf.

For deeper post-sale needs — pre-deployment burn-in, OS / driver / NCCL tuning, networking integration, multi-server cluster setup, private AI stack deployment — solutions are available from the EMARQUE AI team on request. Scope and charges quoted on enquiry. Software / OS / driver / setup support is scoped to the 3-year warranty window.

Step down vs the EMARQUE AI PRO 500 and AI PRO 700: the AI PRO 500 is the Threadripper 5U workstation build (up to 4×) and the AI PRO 700 is the In Win RG650A 6-GPU tier (up to 6×, PCIe Gen4, single 3200W). This AI PRO 800 is the densest RTX PRO 6000 Blackwell box, on the larger IW-RG658-PRO chassis.

Step up vs NVIDIA HGX B300: HGX B300 sits a tier above for frontier training and the highest-throughput inference — 8× B300 Blackwell Ultra with full NVLink GPU-to-GPU interconnect and 2.1 TB HBM3E. This 8× RTX PRO 6000 box is the right tier when full NVLink GPU-to-GPU bandwidth isn't the constraint, and when GDDR7 plus PCIe Gen5 economics fit the workload better than HBM3e plus NVLink.

Step down vs NVIDIA DGX Station: DGX Station is a desk-side single-box workstation (748 GB coherent memory, MIG up to 7 users). This 8× AI Server is production datacentre territory — a different deployment context.

Production AI workloads — the things that exceed a single workstation:

• Multi-tenant inference — serve multiple models or users concurrently with predictable latency. NIM microservices, vLLM, TensorRT-LLM.
• LLM fine-tuning at production scale — SFT, RLHF, DPO, LoRA, QLoRA on mid-to-frontier models.
• Retrieval-augmented generation at organisation scale — embedding pipelines, vector search, document AI for hundreds of thousands of documents.
• Computer vision and multimodal — VLM training and inference, video pipelines.
• Agentic AI with secure tool-calling and autonomous workflows.
• RAPIDS for massive-dataset data science.

Real-world: private GPT-class fine-tuning on enterprise data, organisation-wide internal LLM serving, sovereign-AI deployments where data must stay on-prem.

EMARQUE-built RTX PRO 6000 Blackwell AI Servers ship with Ubuntu LTS plus the full NVIDIA AI Enterprise stack — CUDA, container runtimes, NIM microservices for production inference, AI Workbench, and pre-built framework images (PyTorch, vLLM, RAPIDS, NeMo).

EMARQUE AI Solutions on request can layer in a ready-to-use private LLM, RAG, embeddings and agentic-AI stack pre-configured for your data and use case — chat UI, vector search, document ingestion, model serving, all on-prem. This is the key difference vs a rebranded-chassis integrator: you get NVIDIA's production AI software stack and EMARQUE's deployment support, not just metal with GPUs in it.

Pre-ordering the EMARQUE AI PRO 800 with EMARQUE:

• Locks your allocation for delivery in Malaysia.
• Final pricing confirmed on enquiry — pricing varies by configuration (CPU SKU, system memory, NVMe capacity and tier, networking SKU, software stack, EMARQUE AI services scope). Listed price is an indicative estimate.
• Build, integration and validation lead time confirmed at order — depends on GPU availability and configuration complexity.
• Direct access to the EMARQUE AI team during pre-order for deployment scoping (rack, PDU, networking fabric, multi-server cluster planning, software stack).

Click ENQUIRE NOW above or email business@emarque.co.

EMARQUE is among the first solution providers in Malaysia building multi-GPU AI servers in the RTX PRO 6000 Blackwell tier. What sets EMARQUE apart:

• It is EMARQUE-built. EMARQUE selects, integrates, validates and thermally tunes the system — the In Win 6.5U chassis, PSU, GPU placement, airflow and BIOS — not a relabelled box. Real engineering control over how it runs in your rack, with a direct Malaysia-based warranty path.
• NVIDIA AI Enterprise included. You get NVIDIA's production AI software stack (CUDA, NIM, framework images, AI Workbench) configured and ready — not just metal with GPUs in it. This is the difference vs a rebranded-chassis integrator.
• Solutions available from the EMARQUE AI team — deployment scoping, rack and PDU sizing, ConnectX-8 networking, multi-server clustering, Base Command and Kubernetes orchestration, private LLM / RAG / agent stack deployment, team training. Scope and charges quoted on enquiry.
• Sovereign AI for Malaysian organisations — your data, your models, your stack, on-premise, supported in country.

For buyers comparing AI server providers, the practical question is who can integrate the box, the rack, the network and the AI stack — in country, with one point of contact. Email business@emarque.co for sizing.

Yes. EMARQUE AI is our dedicated AI compute division — full NVIDIA AI hardware stack and on-prem deployment for Malaysian businesses, research labs, and enterprises: NVIDIA Personal AI (DGX Spark, 2× Spark stacking), AI workstations (RTX PRO 6000 Blackwell and others), DGX Station GB300, multi-GPU AI servers, plus full integration support. Visit EMARQUE.AI or contact business@emarque.co.