VM-01
VM-02
VM-03
VM-04
GPU INSTANCES8× H100 SXM · 80 GB
replay-adapter × 4Illustrative loop
01
Ideas, combined.
02
Experiments, parallel.
03
Compute, on demand.
VM-05
VM-06
VM-07
VM-08
EC2 / CLOUD-INITp4d.24xlarge · 8× A100
routing-ablation × 4VM-09
VM-10
VM-11
VM-12
COMPUTE ENGINEa3-highgpu-1g · 1× H100
routing-seed × 4VM-13
VM-14
VM-15
VM-16
VIRTUAL MACHINESNC24ads A100 v4 · 1× A100
held-out-eval × 4POD-17
POD-18
POD-19
POD-20
GPU PODS8× H100 SXM · 80 GB
distill-sweep × 4BOX-21
BOX-22
BOX-23
BOX-24
ELASTIC CONTAINERS4× H100 · GPU sandbox
successor-trial × 4DataCheckpointsEvidence
20 clouds. One system.
Lambda
AWS
Google Cloud
Azure
RunPod
Modal
Crusoe
Cloudflare
DigitalOcean
Vast.ai
Fluidstack
Hyperstack
Paperspace
Vultr
Verda
TensorDock
Thunder Compute
Voltage Park
Akamai / Linode
GiveMeANodetinkerTraining API Planned
Integration coverage
Provider adapters are implemented. Live research workflows have been validated on Lambda A10 and Cloudflare CPU; DigitalOcean and GiveMeANode have CPU transfer validation. Other providers and hardware configurations require live validation. The fleet above illustrates possible workloads, not a running deployment or a performance claim.