runaiicloud
ModelsAppsGPUsServerless GPUPricingReserveDocsConnectCompareEnterprise
Log inGet started
Home/Benchmarks/Park benchmarks

[ benchmarks — live-fired receipts · no simulations ]

No GPU state
ever has to die.

Park freezes a running CUDA workload — process tree, RAM, and VRAM — into one signed artifact. Wake restores it anywhere in seconds, byte-exact. The cold-start chart below is the market's own baseline: loading a 753B-class model from scratch.

Open the Park console →The byte-exactness method ↓
2.3–25 s
wake, by workload class
sha256
identity proof on every restore
$0
idle compute while parked
120 s
spot notice we park inside

Cold start vs restore

The market's headline is “21.7× faster to first token” (61 s restore vs 22 min cold). Here are our receipts against the same 22-minute baseline — per workload class, because a multiplier without its workload class is marketing, not a benchmark.

Cold start (753B-class, disk → VRAM)

22 min

1× baseline

Market restore claim (their number)

61 s

21.7×

runaii Park — engine warm restore (KV cache live in VRAM)

17–21 s

62–77×

runaii Park — container wake

2.3–9 s

146× and up

The engine row is the wedge: park the warm inference engine itself — KV cache and all — instead of replaying conversations to rebuild warmth. Restore ms/GB at torch-image scale: a 5.6 GiB rootfs's first-generation delta backup runs in 17.8 s (~315 MB/s).

The receipt ladder

Every row is a live-fired drill on real GPU hardware — park measured on a loaded workload, wake measured after the process tree was killed and the box wiped. Failed runs are kept in our internal register too; the ladder below is the passing state of the art, one rung per capability.

DrillWorkloadParkWakeProof
D18bdocker-managed GPU container · 1 GiB VRAM canary13 s9 sVRAM sha identical after kill + wipe
D19runc-managed CUDA workload · 2 GiB VRAM16.0 s7.6 s9/9 asserts, canary byte-exact
D19btorch 2.14 image (5.6 GiB rootfs) · 2 GiB torch canary~30 s~25 sallocator pool alive, canary byte-exact
D17cuser session + fs-delta — cross-VM—4.7 s4/4 on a different physical box
D18-crossdocker + GPU container — cross-VM, from object storage alone—24 s8/8, disk wiped on the target
D-skewdump on driver 580.178.04 → wake on 580.173.02——8/8 byte-exact through the older driver
D20REST migrate → wake, same VM (the product path)26.8 s12.5 sworkload counter continues across death
D20 live migrationprefetch + wake on a second machine — the live flip—22.7 s (prefetch 4.3 s)canary byte-exact from the foreign GPU, counter 29→31
D22KERNEL-LAUNCH survival — pre-loaded CUDA modules, same-arch control16 s9 slaunches CONTINUE post-restore (the first kernel receipt — every earlier proof was memcpy-only)
D22H100-native park/wake (Hopper sm90)15 s7 sbyte-exact canary, kernel launches alive on H100
D23CROSS-ARCH migration — park on RTX PRO 6000 (Blackwell sm120), wake on H100 (Hopper sm90), different boxes and driver builds — CR-LITE exec-bridge—6 s17/17: counters CONTINUE across the arch hop (5→10, kernel cell 15→30), canary byte-exact at the original virtual address, fresh cuInit on Hopper
D24MULTI-GPU cross-arch — 2× RTX6000 (g4x2) ⇄ 2× H100 (h100x2), both directions, driver-identical (580.173.02)16 s10 s2/2 children ALL-OK both ways: per-child pid preserved (7→7, 8→8), per-child fixed VA + canary rehydrated, counters continue (k_cell 18→39/15→33 fwd, 15→33/15→30 rev) — plus the PTX up-JIT receipt (sm_90 module loads rc=0 on sm120)
D25THE DOWNGRADE CONSEQUENCE — 32 GiB child vs an occupied 96 GB GPU: fit gate refuses → gate bypassed fails LOUD (bridge OOM, forensics written, artifact intact) → remediated → byte-exact wake203 srefused → forced-fail → PASSGate: 'child 0 holds 32.0 GiB frozen device memory, GPU 0 has only 14.4 GiB free (another process is using this GPU)' → forced: 'bridge cuMemCreate failed: CUresult 2 (out of memory)' → after freeing: counter 5→10, k_cell 15→30, fixed VA, canary rehydrated, 32 GiB re-upload in 1.494 s — plus the TP=8 topology refusal (a single process sharded across 8 GPUs cannot wake on 1)
D26INVOKEAI v5 (the last catalog app) — the w5 lane's first foreign real-app roundtrip, plus a park fired INTO a live generation; spot-reclaimed mid-drill, artifacts outlived the hardware31 s (idle) / 5.3 GiB mid-render34 s (VRAM state refilled: 3015 MiB idle / 3611 MiB mid-render)Post-wake generation PIXEL_DIFF 0 of 786432 — byte-exact vs the parked process's own last render (a process RESTART is only pixel-exact: the wake is strictly stronger than a restart); the mid-generation park's session RESUMED and completed its 1024²/500-step render after the wake; the 2.7 GB model + DB + gallery ride the artifact via the fs-delta bind roots. Five lane walls live-found + fixed (pre-cuInit device-node conversion, mknod umask, docker-exec invisibility to runc checkpoint, the restored-container re-park stdio mount, the secrets-gate doc-literal class)

Spot-preemption rung: park completes in 23.2 s on a 2 GiB workload inside the 120 s preemption notice a spot GPU gives you — the artifact is durable off-box (65.9 s full-durability window at 24 GiB), so the preemption can't kill your state even when it kills the machine.

The byte-exactness method

“Restored” is a claim anyone can type. Every drill above is forced to prove it the same way — the check is the product:

1 · Canary

Before freeze: workload writes known bytes to VRAM at recorded virtual addresses, plus filesystem and process-tree witnesses.

2 · Freeze

Park drains GPU state and the whole process tree into one signed artifact (HMAC over the manifest; untrusted bytes never execute).

3 · Kill

The original is destroyed — process killed, container removed, VRAM wiped. No courtesy state survives.

4 · Wake

Restore on the same box, or a different one from the artifact alone (driver skew allowed down-contract, never up).

5 · Prove

The restored workload re-reads the canary at the original addresses. Byte-exact means sha256 identical — not “approximately resumed.”

What this is not

Not zero-downtime

Hibernation-grade: there is a real park→wake gap (seconds). Nothing on the market does seamless GPU migration either — we publish our gap instead of hiding it.

Cross-arch, receipted both ways

The NVIDIA checkpoint API itself refuses foreign architectures — receipted live: the process tree and files restore, the CUDA context answers “operation not supported” (equally true of anyone riding that API). Our CR-LITE exec-bridge goes around the gate: the workload quiesces to portable bytes, parks under plain CRIU, wakes on the other architecture and rebuilds its context natively — receipted sm120 → sm90 with counters continuing (17/17), the reverse direction too (sm90 → sm120, where the stale sm_90 module loads natively — PTX up-JIT), and multi-GPU (2× RTX6000 ⇄ 2× H100, both directions, per-child pid and counters intact). The honest limit: the bridge is a contract the workload opts into, not a transparent drop-in for arbitrary processes — and the 8× / B200 rungs are receipted as capacity-blocked (quota/stockout), not yet as passing.

Not a sandbox

Park is a state layer: it preserves and moves GPU workloads. Isolation is someone else's race — we can be the state layer underneath it.

Driver floor, not driver indifference

Wake targets must meet a driver-version contract (major-version floor). Skew receipts run down-contract; there is no up-contract magic.

Honest limits, registered

Multi-process CUDA owners that hold the context in a child don't ride yet; sockets with unread buffered data skip+close at freeze; semaphore VALUEs aren't carried (proven non-fatal for vLLM/sglang-class engines).

Park your first GPU workload

One signed artifact. $0 idle. Wake anywhere byte-exact. Try the $0-idle calculator in the console — the math is the pitch.

Open the Park console →Pricing
runaiicloud

Serverless inference, dedicated GPUs, and training for open models. OpenAI- and Anthropic-compatible APIs.

© 2026 runaii

Platform

Model libraryAppsGPUsServerless GPUPricingToken-Max coding plansCompare providersDocs

Developers

ConnectCookbookCLIAgents / MCPUse casesTutorialsBlogFAQ

Company

EnterpriseStartupsAboutCareersTrust centerStatusSupportTermsPrivacyData policySubprocessorsSitemap