01
Measure before keeping
No prediction, heuristic or model output is ever kept without a measurement on the target GPU.
Security and correctness
Turbokern changes the code that runs your models. These are the rules it follows so that every kept change is correct, accurate within your budget, and traceable.
Principles
01
No prediction, heuristic or model output is ever kept without a measurement on the target GPU.
02
A kernel is benchmarked only after it passes a numerical equivalence check.
03
Accuracy is judged on your evaluation set against the budget you set.
04
Every candidate, result and decision is written to the run report.
Method
Kernel changes are compared to the reference implementation on representative and edge-case inputs, within a numerical tolerance per data type.
Precision and sparsity changes are scored on your evaluation set. Each score carries a confidence interval, and results the eval set is too small to resolve are flagged rather than kept.
Example run on one H100. Candidates, speedups and reasons are illustrative, not measured results.
Standards context
Turbokern can produce evidence that supports your own compliance work. It is not a certification body and holds no certifications today.
| Framework | What it asks | How Turbokern can help | Status |
|---|---|---|---|
| SOC 2 | Controls over security and availability | Turbokern's own SOC 2 Type II audit | Planned |
| ISO/IEC 27001 | An information security management system | Policies designed to align; certification not yet sought | Not started |
| EU AI Act | Documentation and quality management for some AI systems | Run reports document model changes and their measured effect | Evidence only |
| ISO 26262, IEC 62304 | Functional safety for automotive and medical software | On-device measurements for your validation; no safety claims | Evidence only |
Audit trail
The run report is an append-only log. Each entry records the candidate, the hardware, the software versions, the measurement and the verdict. Reports can be exported to your tracking system.
{"id": "C-067", "stage": "quantra",
"gpu": "H100-SXM-80GB", "cuda": "12.6",
"p95_ms": 41.8, "acc_delta": -0.003,
"verdict": "kept"}
// example entry, illustrative valuesSecurity
FAQ
In the managed mode, yes, inside your isolated tenant. In VPC and on-prem modes the weights never leave your infrastructure.
No. A SOC 2 Type II audit is planned before general availability and has not started.
Within the tolerances and accuracy budget you set, yes; that is what optimization means. Anything outside them is rejected and logged.
You are. Turbokern packages and documents it; by default a person on your team approves deployment.
Next step
Our security contact will answer questions before you share any model.