GPU Workstations

Overview

Remangu GPU Workstations deliver enterprise-grade rendering and simulation performance in a fully managed cloud environment. Each workstation is backed by NVIDIA RTX 4090, A6000, or H100 GPUs and can be provisioned with up to 512GB of system RAM and 8TB of NVMe storage. Artists, technical directors, and engineers connect through the NICE DCV streaming protocol, which provides pixel-perfect fidelity at up to 4K resolution with sub-30ms latency on supported networks.

Every instance ships pre-configured with licensed installations of Unreal Engine, Unity, Autodesk Maya, and Blender, eliminating the multi-day setup cycles that typically precede production work. Infrastructure scales on demand: spin up additional workstations during crunch, release them when the milestone ships, and pay only for the hours consumed.

Key Features

• High-Performance GPU Acceleration — Select from NVIDIA RTX 4090 for real-time ray tracing, A6000 for large-scene visualization with 48GB VRAM, or H100 for AI-assisted content generation workflows.
• Pre-Configured Creative Tools — Workstations boot with production-ready installations of Unreal Engine 5, Unity 6, Maya 2025, and Blender 4.x. Plugin ecosystems and studio-specific toolchains are preserved across sessions.
• NICE DCV Streaming Protocol — Adaptive streaming adjusts bitrate and codec selection based on network conditions, supporting multi-monitor layouts, stylus input, and USB device redirection.
• Auto-Scaling — Workstation pools expand and contract automatically based on team schedules or CI/CD triggers, ensuring resources are available when artists need them without idle cost accumulation.
• Dedicated Persistent Storage — Each workstation mounts dedicated NVMe volumes that persist across sessions, with optional shared project storage accessible to the entire team via high-throughput NFS.

Technical Specifications

Specification	Detail
GPU Options	NVIDIA RTX 4090 (24GB) / A6000 (48GB) / H100 (80GB)
System RAM	64GB to 512GB DDR5
Storage	256GB to 8TB NVMe SSD per instance
Streaming Protocol	NICE DCV with H.264/H.265 encoding
Network	Up to 100 Gbps instance bandwidth
OS	Amazon Linux 2, Ubuntu 22.04, Windows Server 2022

How It Works

1. Provision — Select a GPU tier, RAM configuration, and storage capacity through the Remangu console or Infrastructure as Code templates. Workstations are ready in under five minutes.
2. Configure — Apply a studio profile that installs your project’s engine version, plugins, and pipeline tools automatically. Profiles are version-controlled and shareable across teams.
3. Connect — Artists launch the NICE DCV client on any device—Windows, macOS, Linux, or browser—and authenticate through your identity provider. Sessions resume exactly where they left off.
4. Scale — As your team grows or deadlines compress, add workstations from the pool. When utilization drops, instances hibernate or terminate based on policies you define, keeping costs aligned with actual usage.