CASE STUDY · HEALTHCARE · NORTHSTAR MEDICAL RADIOISOTOPES
How NorthStar Medical Radioisotopes Cut Simulations from 4 Days to Under an Hour
✓96× faster Monte Carlo N-Particle (MCNP) simulations, same models and inputs
✓Moved off aging on-prem hardware to a TurbOS® private cloud in one day
✓33% more research throughput, with simulations rising from 60% to 80% of engineer time
✓50% lower cost than standard public cloud, with no added staff or IT burden
Multi-day simulations were the bottleneck. TurbOS removed it.
- Why multi-day MCNP runs were holding back NorthStar's product and regulatory timelines
- The four points of friction: cost to scale, slow results, aging capacity, and an HPC expertise gap
- How TurbOS maps MCNP's parallel model to optimized compilers, scheduling, and dedicated private cloud nodes
- The five-step timeline that took NorthStar from legacy on-prem to live in a single day
- Measured outcomes: 96× faster runs, 33% more throughput, 50% lower cost than public cloud
- Why Hoonify is one of the only private operators endorsed by Oak Ridge National Laboratory's RSICC to host MCNP
Running MCNP or other simulation workloads? Request a TurbOS® demo
Download the NorthStar case study
THE CHALLENGE
Days of waiting on every run
NorthStar needed fast, repeatable MCNP results to keep radiopharmaceutical programs on schedule. Older on-prem clusters were out of space and power, the IT team lacked deep HPC tuning experience, and public cloud brought unpredictable pricing and compliance risk.
Fewer design points per week meant slower research and fewer patients helped.
THE RESULT
“It ran so fast I thought it was broken.”
Same models, same inputs, no change to how engineers work. Hoonify delivered TurbOS as a turn-key private cloud, so MCNP jobs that took four days now finish in under an hour.
Josh Peterson-Droogh, Nuclear Engineering & Physics Manager, NorthStar