
==============================================================================
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=8)
  files      : 20489
  on disk    : 53.18 GB

Reading the RAMSES output (16 threads):
  hydro      :    53.17 s +- 2.00 s
  particles  :     3.70 s +- 0.89 s
  gravity    :    31.07 s +- 0.73 s
  TOTAL      :    87.94 s

Conversion (hydro, gravity, particles, 16 threads):
  read from RAMSES   :    83.50 s
  savedata write     :     5.58 s
  one-off conversion :    89.09 s
  MERA re-read warm  :   1.87 s   (44.7x faster)
  size on disk       : 53.18 GB -> 819.3 MB   (levels up to 8 only, not comparable)
  break-even         : 1.1 re-reads
  memory to load the same data:
    allocated RAMSES :  110.33 GB
    allocated MERA   :    2.96 GB   (37.3x less churn)
    GC RAMSES / MERA : 16.31 s / 294.0 ms
    peak RSS RAMSES  :   10.36 GB
    peak RSS MERA    :    7.16 GB   (1.4x 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 :   204.39 s  1.00x
    2 threads :   120.39 s  1.70x
    4 threads :    80.97 s  2.52x
    8 threads :    60.48 s  3.38x
   16 threads :    53.17 s  3.84x
   24 threads :    56.03 s  3.65x
  fastest 16 threads, sweet spot 16 threads

Storage, IOPS by thread count:
    1 threads :     105900 IOPS
    2 threads :     193479 IOPS
    4 threads :     247571 IOPS
    8 threads :     274401 IOPS
   16 threads :     302206 IOPS
   24 threads :     280053 IOPS
==============================================================================
