RTX 4090      x1     OB  1330   ████████░░░░░░░░░░░░  local
RTX 4090      x3     OB  3990   ████████████████████  EVA · NODE_07
RTX 5080      x4     OB  3976   ████████████████████  EVA · NODE_03
RTX 5090      x2     OB  4180   ████████████████████  EVA · NODE_11
─────────────────────────────────────────────────────────────────────
INPUT      frames=240  avg_min=8            local_total =   1920 min
SOLVE      throughput = local / eva         eva_total   =    640 min
COST       eva_min × rate × nodes           ≈ 4 270 ₽   ·  $58.7
SPEEDUP    local / eva                      = 3.0×        ↑↑↑
SAVED      local − eva                      = 21.3 h

calculator · by OctaneBench

We price rendering. Honestly, by benchmarks.

Pick your configuration and see what the same job would cost on the Eva Cluster. We calculate via OctaneBench — the industry benchmark by Otoy, with no marketing multipliers. Minimum rental is 1 hour, then per-minute billing, in rubles or dollars.

OctaneBench, not vendor “fps” claims 4090 / 5080 / 5090 — the current fleet RUB and USD, per-minute after the first hour

01 · frames1,920 GPU-minutes of work

pcs

15000

02 · min per frame

min

1120

03 · your hardwarewhat you have now

GPU

cards

power · OctaneBench1,330

your render time

1 d 8 h

04 · Eva Clusterwhat we offer

node type

nodes

power · OctaneBench3,990

time on Eva

10 h 40 min

05 · task cost on Eva

4,267 ₽

1 node × 10 h 40 min

rate 400 ₽/hminimum 1 hourthen per-minute

faster by

3.0×

saved

21h

✦

Speed is computed via OctaneBench 2025.2 by Otoy — no marketing multipliers. Minimum rental is 1 hour, then billing is per-minute. This is an upper-bound estimate — in practice jobs often run faster.