
==============================================================================
MERA BENCHMARK REPORT
==============================================================================
Date         : 2026-09-07
Host         : op4pgn
CPU          : INTEL(R) XEON(R) GOLD 6534 (32 threads)
RAM          : 1511.41 GB
Filesystem   : btrfs   mount /data1
Julia        : 1.12.7, 24 compute + 24 GC threads
Mera         : 1.8.0

Dataset      : L13_SN5_CD_only output 390
  ncpu       : 5120
  levelmax   : 13   (read capped at lmax=7)
  files      : 20489
  on disk    : 53.18 GB

Reading the RAMSES output (16 threads):
  hydro      :    31.84 s +- 0.40 s
  particles  :     4.60 s +- 1.46 s
  gravity    :    26.90 s +- 1.25 s
  TOTAL      :    63.34 s

Conversion (hydro, gravity, particles, 16 threads):
  read from RAMSES   :    61.86 s
  savedata write     :     3.79 s
  one-off conversion :    65.65 s
  MERA re-read warm  : 270.8 ms   (228.4x faster)
  size on disk       : 53.18 GB -> 179.9 MB   (levels up to 7 only, not comparable)
  break-even         : 1.1 re-reads
  memory to load the same data:
    allocated RAMSES :    75.7 GB
    allocated MERA   :   612.3 MB   (126.6x less churn)
    GC RAMSES / MERA : 15.36 s / 50.2 ms
    peak RSS RAMSES  :    3.66 GB
    peak RSS MERA    :    3.31 GB   (1.1x lower)

Storage by component:
  hydro        24.94 GB   46.9%
  gravity      17.52 GB   32.9%
  amr          10.66 GB   20.1%
  particle      65.0 MB    0.1%

Reading thread sweep (:hydro):
    1 threads :   155.34 s  1.00x
    2 threads :    83.36 s  1.86x
    4 threads :    55.12 s  2.82x
    8 threads :    40.54 s  3.83x
   16 threads :    32.06 s  4.84x
   24 threads :    31.46 s  4.94x
  fastest 24 threads, sweet spot 16 threads

Storage, IOPS by thread count:
    1 threads :      99185 IOPS
    2 threads :     184638 IOPS
    4 threads :     229921 IOPS
    8 threads :     245544 IOPS
   16 threads :     303129 IOPS
   24 threads :     294936 IOPS
==============================================================================
