
==============================================================================
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=11)
  files      : 20489
  on disk    : 53.18 GB

Reading the RAMSES output (16 threads):
  hydro      :   144.38 s +- 9.91 s
  particles  :     3.38 s +- 1.03 s
  gravity    :    55.96 s +- 4.37 s
  TOTAL      :   203.73 s

Conversion (hydro, gravity, particles, 16 threads):
  read from RAMSES   :   203.31 s
  savedata write     :     9.67 s
  one-off conversion :   212.98 s
  MERA re-read warm  :   6.17 s   (32.9x faster)
  size on disk       : 53.18 GB -> 3.48 GB   (levels up to 11 only, not comparable)
  break-even         : 1.1 re-reads
  memory to load the same data:
    allocated RAMSES :   250.3 GB
    allocated MERA   :   13.03 GB   (19.2x less churn)
    GC RAMSES / MERA : 32.74 s / 492.8 ms
    peak RSS RAMSES  :   31.11 GB
    peak RSS MERA    :   26.53 GB   (1.2x 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 :   365.26 s  1.00x
    2 threads :   240.92 s  1.52x
    4 threads :   183.11 s  1.99x
    8 threads :   155.64 s  2.35x
   16 threads :   145.48 s  2.51x
   24 threads :   149.87 s  2.44x
  fastest 16 threads, sweet spot 16 threads

Storage, IOPS by thread count:
    1 threads :     105173 IOPS
    2 threads :     181979 IOPS
    4 threads :     228708 IOPS
    8 threads :     270090 IOPS
   16 threads :     288500 IOPS
   24 threads :     292794 IOPS
==============================================================================
