PROFESSIONAL GPU COMPUTE
PRO 6000 Servers
GPU server options for AI development, inference and GPU-accelerated creative workloads.
Professional GPU compute
Build, test and create.
AI experimentation and creative workloads place different demands on a server. Start with the software you use, the size of your data and the memory needed by your model.
Discuss PRO 6000 server options for prototyping, application inference and content generation. The exact accelerator variant and complete server configuration are confirmed in your proposal.
WORKLOAD & CONFIGURATION
Match the server
to the workload.
Discuss your software stack, dataset size and accelerator memory requirements with NGT. The exact PRO 6000 variant and server specification are confirmed in your proposal.
Workloads to discuss
- AI prototyping and experimentation
- Image and content generation
- Application inference
- Model fit
- Model size, numerical precision, framework and accelerator memory.
- System resources
- CPU, system memory and data preparation requirements.
- Data access
- Dataset storage, networking and expected data movement.
- Rental plan
- U.S. location, target start date, monthly or annual term.
Product images are references. Exact GPU variant, complete server configuration, location availability and commercial terms are confirmed in your proposal.
Useful details to share
- Your model, software stack and workload duration
- Memory, storage and connectivity requirements
- Preferred U.S. location and rental term