Estimate the number of GPUs needed for peak concurrent AI transcription workloads. The model combines measured real-time factor, batching benefit, downstream GPU load, usable-capacity headroom, and redundancy. It reports both the fractional infrastructure requirement and the whole-GPU provisioning count, helping teams avoid capacity estimates based only on raw concurrency.
Raw GPU load = concurrent streams × real-time factor × (1 − batching gain) × (1 + post-processing load). Required GPUs = ceiling(raw load ÷ usable GPU capacity × replica multiplier).
The main result is a planning estimate for one transcription workload or its available capacity. It is only as accurate as the pricing, model-speed, token-density, and workload assumptions entered. Use measured production values whenever possible and run high- and low-case scenarios before committing capacity or a customer price.
This planning estimate does not guarantee vendor billing, model performance, transcription accuracy, translation quality, or service-level objectives.
At 40 streams, 0.18 RTF, 15% batching gain, 10% extra load, 80% usable capacity, and two replicas, the adjusted requirement is 16.83, so provision 17 GPUs.
Use representative production values or a weighted average from recent tasks. Peak values are better when testing capacity risk.
Not automatically. Increase workload, overhead, tokens, or utilization headroom to model retries and operational variance.
Yes. Enter each option’s measured speed, infrastructure assumptions, and labor or usage prices as separate scenarios.
Headroom helps absorb variable media length, traffic bursts, startup time, retries, and queueing without immediately missing targets.
Update it after pricing, models, hardware, workflow stages, average task size, or concurrency patterns change.
| Variable | Meaning |
|---|---|
| Workload | Media length, source volume, or concurrent demand entered above. |
| Efficiency | Measured model rate, token density, batching gain, or utilization assumption. |
| Overhead | Workflow work outside the primary model calculation. |
| Result | Estimated requirement or capacity before unmodeled operational variance. |