AI Inference
Submit prompts to supported models through the AI inference experience.
Explore AI inference ↗SOLARCOMPUTE / WORKLOADS
Real work.
Distributed compute.
From AI inference and rendering to training, simulations and CPU jobs, explore workload pathways on SolarCompute's distributed computing network.
COMPUTE SERVICES
Choose the right
workload.
Six pathways into SolarCompute. Each workload has its own requirements, controls and eligible machine configurations.
AI Training
Explore controlled training and fine-tuning workloads on approved machine configurations.
Explore training ↗Rendering
Explore image generation and GPU-assisted rendering on compatible resources.
Explore rendering ↗Simulation
Explore numerical, modelling and scientific compute with approved runtime requirements.
Explore simulation ↗CPU Jobs
Run supported calculations, batch processing and non-GPU tasks.
Explore CPU jobs ↗Custom Containers
Explore approved container-based workloads with platform controls and suitable hardware.
Explore containers ↗HOW IT WORKS
Hardware determines
the workload.
SolarCompute's Hardware Matrix governs which workload a specific hardware configuration is approved to execute. Minimum specifications alone do not authorise a workload.
01Discover hardware and normalise its profile
02Check the exact configuration against the Hardware Matrix
03Apply the approved workload recipe and runtime
04Execute on eligible capacity and return job results
PLATFORM EXPERIENCE
From submission
to results.
Explore a workload service, submit an eligible request and follow its progress through the supported application or dashboard.
Public demonstrations and production use may have different permissions, availability, resource limits and pricing. Refer to the individual workload page for the applicable experience.
WORKLOAD LIFECYCLE
Submitted → Queued → Assigned → Running → Results
Execution is subject to approved machine capability, network availability and workload-specific controls.
CONTROLLED EXECUTION
Built for
responsible use.
Workloads need appropriate permissions, resource limits and execution environments. Supported controls vary by service.
Approved recipes
Only authorised workload definitions and runtimes should execute.
Resource limits
Runtime and capacity constraints protect network resources.
Access controls
Public, private and advanced workflows have different requirements.
Job visibility
Track supported execution states and retrieve permitted results.
START EXPLORING
Find your next
workload.
Choose a service or learn how the distributed network supports approved compute jobs.