AiSilicon provides an innovative workflow on GPU hardware, priced hourly.
Run as many compute jobs as you need simultaneously on dedicated NVIDIA GPUs in a Tier 4 Datacenter. We can scale to meet your needs.
No paying for idle resources here.
We only bill for the time your jobs are running, plus competitive prices for dataset storage.
Our platform provides a powerful workflow and collaborative tools that enable teams to reign in costs and work together effectively.
We designed a workflow that provides a simple and flexible approach to provisioning and pricing.
And then we built a platform for it.Create a project and get started
How does it work?
Create a project on AiSilicon
Push a branch to your new Git remote
Pull results after your job completes
Simply push your codebase and any data to the Git+LFS repository we provide and start running your project right away.
Every job runs in a Docker container with configurable dedicated GPUs, giving you the flexibility to run any workload.
Pushing a branch to your git repository creates a job automatically. Once your program exits, billing stops immediately.
We only charge these rates while your jobs are actively running.
6144 CUDA cores, 192 Tensor cores, 16GB VRAM
4 dedicated Xeon CPU cores
30GB system memory
10,752 CUDA cores, 336 Tensor cores, 48GB VRAM
8 dedicated Xeon CPU cores
60GB system memory
Scale your compute resources to meet your needs, without wasting money.