Compute
Bare-metal and virtual instances that boot in seconds and bill by the second — with committed-use discounts that actually compound.
$0.0071 / vCPU-hr

profilo-studio
Made by ariya-tech.net
Singapore · 14 regions · Since 2016
Compute, object storage, managed Kubernetes and a global edge network — laid out on a single lattice of regions engineered for predictable latency and uptime you can build a business on.
All systems operational
99.99%
Uptime SLA
14
Regions
<40ms
Inter-region p95
2.4M
vCPUs online
Every service runs on the same private network fabric, so your systems talk to each other at rack latency — not internet latency.
Bare-metal and virtual instances that boot in seconds and bill by the second — with committed-use discounts that actually compound.
$0.0071 / vCPU-hr
S3-compatible buckets with flat egress and eleven nines of durability. No request-count roulette waiting on your monthly invoice.
$0.011 / GB-mo
Conformant clusters with a control plane we run, patch and page ourselves for. You get kubectl; we get woken up.
control plane free
210 Tbps of anycast capacity terminating TLS and caching at 63 points of presence — one BGP announcement from your users.
63 PoPs · 210 Tbps
Fourteen regions stitched together over private backbone. Traffic between them never touches the public internet — and never pays egress to move.
Anycast edge, to a health-checked load balancer, to a pool of stateless nodes, to replicated storage — the same path whether you serve ten requests or ten million.
Anycast edge
TLS + cache
Load balancer
health-checked
Compute pool
stateless · autoscaled
Replicated storage
3× across zones
We moved forty services onto the lattice in a weekend. Our p99 dropped by half, and so did the invoice.
We will tell you what it costs on the lattice — and where it would break — before you move a single byte.
Start building