docs(nvme): record the ordered MP600 Mini 2TB and its acceptance test

2026-11-04: Corsair MP600 Mini 2TB ordered for dgx4. Class: Phison PS5027-E27T
(4-channel, DRAM-less), 64 MB HMB, 162L Kioxia BiCS6 TLC, 1200 TBW. Same
architecture family as the KingSpec, but a newer controller with a larger HMB
- and the healthy reference drive in this fleet is also HMB-based, so
DRAM-less alone is not the verdict: only the measurement decides.

- Corrects the earlier ranking: MP600 Micro (2242) is the OLDER Phison E21T
  design, a controller generation behind the Mini's E27T.
- Pre-registers the acceptance gates (98 GB random at 16/64 threads, then the
  two engine tests on prod4b: cold 157K prompt + full bench) so the drive is
  judged by numbers, not by a hunch.
- Mounting note: 2230 mounts in the GX10's 2242 slot on the single screw
  (owner-confirmed); adapter bracket + thermal pad as fallbacks.

specs/tech-architecture/dgx4-nvme-investigation.md

@@ -97,19 +97,53 @@ helps on the warm/cached path and does not save the cold or sustained cases.
 
 ## What to buy instead (if dgx4 must serve)
 
-The requirement is sustained random-read at a large working set, i.e. a
-**DRAM'd controller with a big FTL map cache**, not peak sequential speed:
+The requirement is sustained random-read at a large working set.
+
+**DRAM-less alone is not the verdict.** The healthy reference drive in this
+fleet (`ESL01TBTLCZ`) *also* declares a host memory buffer (64 MB) and stays
+flat across working sets — what separates it from the KingSpec is the FTL's
+map-cache behaviour and HMB implementation, not the mere absence of DRAM.
+
+Ordered 2026-11-04: **Corsair MP600 Mini 2 TB** (M.2 2230). Class per
+TechPowerUp/KitGuru: Phison **PS5027-E27T** (4-channel, DRAM-less), **64 MB
+HMB**, 162-layer Kioxia BiCS6 **TLC**, 1,200 TBW, 5-year warranty — the same
+architecture family as the problem drive, but a newer and much better regarded
+controller with a larger HMB. Worth measuring, not assuming.
+
+Naming trap: the **MP600 Micro** (2242) is the *older* Phison **E21T** +
+Micron 176L design — a controller generation behind the Mini's E27T. The
+shorter model name is not the better drive.
+
+Alternatives if the Mini fails the acceptance test:
 
 1. **Samsung PM9A3 M.2 2242** (enterprise, Elpis 8-channel, DRAM) or
    **PM9E1 2242** if reachable — designed for exactly this duty.
 2. **Corsair MP700 Micro 2 TB** (Gen5 2242, Phison E28, 8-channel with DRAM).
-3. **Corsair MP600 Micro 2 TB** (2242, Gen4) — consumer DRAM-equipped class,
-   the floor of what is acceptable.
-4. **Avoid** anything DRAM-less/HMB in this role (XG7000, MP600 Mini 2230,
-   WD SN770M): they are the same architecture as the problem drive.
+3. Avoid a small HMB plus a budget FTL in this role (XG7000, WD SN770M).
+
+### Acceptance test (run before trusting the drive with a lane)
+
+Same harness as below, then the engine. A candidate must clear both:
+
+| gate | threshold (healthy class) |
+|---|---|
+| 4 K random in the 98 GB shard, 16 threads | **>= 60 K IOPS and p50 <= 250 us** |
+| same, 64 threads | **>= 120 K IOPS, p99.9 <= 2 ms** (no congestion collapse) |
+| sequential (`dd bs=1M iflag=direct`) | >= 2.5 GB/s (sanity, not the point) |
+| **engine**: cold-boot 157 K first-touch prompt on `prod4b` | completes, no poison |
+| **engine**: full 97-scenario bench on the same boot | completes, no poison |
+
+The two engine rows are the real test — the harness only predicts them.
 
 Before any reinstall: back up `~/.config/flare/orbit/controller.key` — restoring
-it after a DGX OS reimage avoids re-pairing all eight agents.
+it after a DGX OS reimage avoids re-pairing all eight agents. The package can
+be re-sourced from dgx1's agent cache (`NODE:/...`) or re-downloaded (~50 min)
+rather than copied through the Mac.
+
+Mounting: the GX10 slot is M.2 2242, and a 2230 module mounts on the same
+single screw (an ASUS Ascent GX10 owner confirmed 2230 support); a 2230->2242
+adapter bracket is the fallback, and a thermal pad is worth adding since a 2230
+has less surface area than the stock 2242.
 
 ## Reproducing the measurement