GamerTech
Custom | AMD Threadripper Pro 9995WX | 2× NVIDIA RTX Pro 6000 | Workstation
GamerTech
Custom | AMD Threadripper Pro 9995WX | 2× NVIDIA RTX Pro 6000 | Workstation
Custom | AMD Threadripper Pro 9995WX | 2× NVIDIA RTX Pro 6000 | Workstation
GamerTech Head Office
470 North Rivermede Road
Vaughan ON L4K 3R8
Canada
Best for: AI/ML training labs, large VFX studios, multi-GPU rendering pipelines, and research environments needing 192GB of total GPU VRAM in a single hand-built workstation.
192GB of total GPU VRAM. 96 Threadripper Pro cores. 512GB of ECC system RAM. This is the dual-GPU configuration of our flagship Pro 6000 build — designed for AI training, large-scale GPU rendering, and any pipeline that benefits from parallel GPU compute. Dual Pro 6000s deliver 384 fp32 TFLOPs of compute, run two model training jobs in parallel, accelerate Octane/Redshift/V-Ray GPU renders by 2×, and enable workflows that simply cannot fit on a single GPU. Built around the WRX90 Pro platform with a 1650W Titanium PSU for sustained dual-GPU load.
What's Inside
| Component | Specification |
|---|---|
| CPU | AMD Ryzen Threadripper Pro 9995WX — 96 Cores / 192 Threads, Zen 5 Architecture |
| GPU | 2× PNY NVIDIA RTX Pro 6000 Blackwell — 192GB Total GDDR7 |
| CPU Cooler | be quiet! Silent Loop 3 420mm AIO Liquid Cooler |
| RAM | 512GB (4×128GB) DDR5-5600 ECC Registered |
| Storage | 2× 4TB Gen5 M.2 NVMe SSD (8TB total) |
| Motherboard | ASUS Pro WS WRX90E-SAGE SE — Pro Platform, 8 DIMM, PCIe 5.0 |
| PSU | 1650W 80+ Titanium Fully Modular |
| OS | Configurable — Windows 11 Pro for Workstations, Windows Server, or Linux |
| Warranty | 1 Year Parts + Labour (GamerTech) |
| Shipping | White-glove delivery available. Free standard shipping Canada-wide |
Dual GPU Changes the Game
Two Pro 6000 GPUs aren't just twice the speed — they're two different workloads at full speed simultaneously. Train a model on GPU 0 while running inference for production on GPU 1. Render the final frame on one card while the artist iterates on the next shot on the other. Process two video timelines in parallel. Run multiple LLMs concurrently for RAG and reasoning pipelines. With 192GB total GDDR7 VRAM, you can hold the largest open-weights models, train at higher batch sizes, or run massive 3D scenes that would never fit on a single card. The 96-core CPU and 1650W PSU ensure both GPUs run at sustained full load.
Who Buys This
This workstation is for: AI/ML training and inference labs running production pipelines. VFX studios with parallel GPU rendering. Research labs running multi-model experimentation. Production studios with seat-based render farms. Companies serving AI/ML inference internally with redundancy. Engineering firms running multi-GPU CFD or visualization workloads.
Not the right fit if: Workloads that don't parallelize across multiple GPUs — single-GPU users would be better served by our $62,999 single Pro 6000 build. For triple GPU configurations, see our $99,999 triple Pro 6000 Max-Q workstation.
Workload Performance
| Workload | Performance |
|---|---|
| AI Training (parallel GPUs) | Industry-leading — 192GB VRAM enables larger batches |
| Multi-Model Inference | Excellent — run separate models per GPU |
| GPU Rendering (Octane/Redshift, multi-GPU) | Industry-leading — near-2× single-GPU speed |
| Local Multi-LLM (70B+ + RAG models) | Excellent — both GPUs working in parallel |
| Stable Diffusion Production Pipeline | Excellent — 1 training, 1 inference simultaneously |
| VFX Rendering (CPU + dual GPU) | Industry-leading — full pipeline at top speed |
| Scientific Computing (multi-GPU CUDA) | Excellent — 192GB VRAM + 96 cores |
| 8K Multi-Stream Production | Excellent — separate streams per GPU |
Buying Guides & Resources
Every Build. Every Time.
Frequently Asked Questions
Can this workstation be customized?
Yes. We offer triple GPU configurations (3× Pro 6000 Max-Q) at $99,999 for environments needing 288GB total VRAM. We can also scale RAM to 1TB+, configure NVMe RAID, or add 10/25/100GbE for AI cluster integration. Call (905) 247-7085.
Why dual GPU instead of one bigger system?
Two GPUs let you run two workloads simultaneously — train one model while serving inference for another, or render one frame while iterating on the next. For training large models with batch sizes that exceed 96GB single-GPU memory, two GPUs allow batch sizes up to 192GB combined. For VFX, dual GPU rendering nearly doubles your output per workstation.
Does software automatically use both GPUs?
For most modern AI/ML frameworks (PyTorch, TensorFlow, vLLM, Ollama) — yes, multi-GPU is well-supported and often automatic. For rendering, software like Octane, Redshift, V-Ray GPU, and Blender Cycles all support multi-GPU. For Adobe and DaVinci, multi-GPU support varies by feature. We can pre-configure software for your specific multi-GPU workflow on request.
How long does delivery take?
Estimated 4–6 weeks from order. Dual Pro 6000 configurations require careful component sourcing — contact us for current allocations and white-glove delivery options.