Workload Launch Wizard

Submit Container Workload

Deploy your containerized training, inference, or batch processing pipeline with multi-objective scheduler placement.

Quick Start Workload Presets

PyTorch 2.2 CUDA

1x GPU · 24GB VRAM

NVIDIA CUDA 12.2 Base

1x GPU · 16GB VRAM

HuggingFace Transformers

2x GPU · 48GB VRAM

FFmpeg GPU Transcoder

1x GPU · 8GB VRAM

OCI / Docker Container Image *

Execution Command / Script

Multi-Objective Scheduler Strategy

Balanced (Optimal)

Best price/speed ratio

Lowest Cost ($)

Maximizes budget savings

Max Speed (TFLOPS)

Fastest enterprise GPUs

Max Uptime (99.9%)

Highest node availability

Custom Weights

Fine-tune objective ratios

Dedicated GPU Accelerators

Minimum VRAM (GB)

Environment Variables