Whether MetroVolt's core holds heat comes down to microturbulence. The series builds its case in tiers: reduced-order gyrofluid physics computed everywhere, and full nonlinear gyrokinetic decks (CGYRO) frozen, documented, and waiting on supercomputer time.
Tier-1 analyses use TGLF-class quasi-linear transport with the frozen profiles: they identify the ion-temperature-gradient branch as the governing mode, show the near-thermal baseline (Ti/Te → 1) raising the ITG threshold ≈1.3× over the hot-ion case, and find density peaking at Greenwald fraction ≈1.15 stabilizing the trapped-electron branch. The series is explicit that TGLF is not a nonlinear simulation — that distinction is written into the register.
Tier-2 is the test: deposited CGYRO input decks (S13) with acceptance criteria pre-registered, so when HPC allocations run them, pass/fail was defined before the answer existed.
Confinement is our named bet, and this is its audit trail. Pre-registered acceptance criteria mean the community — not our marketing — grades the result. That structure converts a physics risk into a scheduled, third-party-checkable milestone.
| Governing instability | ITG branch (TGLF-class, Tier-1) |
| Near-thermal advantage | ITG threshold ≈1.3× higher at Ti/Te → 1 |
| TEM stabilizer | density peaking at f_GW ≈ 1.15 |
| Tier-2 test | nonlinear CGYRO decks (S13), pre-registered |
| Status | decks deposited, pending HPC execution |