Played sessions
Testing is done with Walnut's rating tool (walnut-rate, on the console): how ratings are made and submitted.
Builds:
Walnut 0.1,
armagetronad 0.2.9.3.0-1+ps3.816
The platform builds under it (3)
| Linux kernel | 6.12.111+ps3.747 Walnut's build |
|---|---|
| Mesa | 26.2.4-1+ps3.721 Walnut's build |
| ps3rsx | 0.735+k6.12.111+ps3.747 Walnut's build |
| Date | Launcher | Rules | Play | Ended | Picture | |
|---|---|---|---|---|---|---|
| 2026-10-25 | Armagetron Advanced | 1.0 | 9 min | quit | ok | Details |
Download all sessions as CSV (one row per session, the aspects as columns)
Rating history
The rating on each version it was tested on, newest first: each row is rated from that version's sessions alone, under the newest rules those sessions have the data for.
| Version | Tested | Medal | Rules | Frame rate | Smoothness | Sound | Picture | Stability | Display | Load | Play |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Walnut 0.1 armagetronad 0.2.9.3.0-1 |
2026-10-25 | Platinum · 95 | 1.0 | 100 | 84 | 97 | 100 | 90 | - | 100 | 0.1 h 1 test |
Scripted benchmarks
Not rated: tests of driver and game changes. A script plays the same scene on the PS3 again and again (the desktop stopped, vsync off) to compare two builds of the game or of the graphics driver. The numbers say which build is faster; they never feed the medal.
How the script plays it: Armagetron Advanced 0.2.9.3: a local game of 4 AI players with the player spectating, which plays itself round after round. fps from the Gallium HUD.
The recipe's full description
Armagetron Advanced 0.2.9.3: a local game of 4 AI players with the player spectating, which plays itself round after round (SPECTATOR_MODE_1 and the SP_* settings make it AI only; setup writes them for every run). fps from the Gallium HUD. Menus (0.2.9: FIRST_USE is true on a first start, so FIRST_USE 1 in the config means the first start's language menu, player setup and welcome screen, which wait for a key: the run of 2026-10-18 sat at the language menu for its whole window). The live variants write FIRST_USE 0 and the language (LANGUAGE_FIRST/_SECOND), and `during` types Return, Down, Return into a virtual keyboard (as etr's recipe does): main menu "Play Game", then "Local Game", which starts the game at once. Playback and record keep FIRST_USE 1: their recording holds the first start's keys. Steady state: the HUD window (WINDOW_STOP=1: 60 s from the first frame, then 120 s, then the game is stopped). The rounds are alike but not identical from run to run (the AIs decide live). The match limits are raised (10 rounds or 30 minutes by default: the champion screen and a new match every few minutes otherwise). Until 2026-10-18 base played back assets/armagetronad-bench.aarec (--playback --benchmark), a 90 s recording cut by the record run's stop, not ended with a quit: after its end the game draws nothing more (a black screen with the cursor, no frames, no CPU: the HUD stopped 66-77 s after its first frame in every run), so the 60 + 120 s window held ~6 s of frames and perf had samples only at its start (the suite run 20261018T004040Z). The recording is variant playback now, with a window that fits in it; a longer one (variant record, ended with Quit from the menu before DURATION) would make a playback window of the full length.
9 variants of the recipe
base stands for the game | the recipe as it is, without a variant |
|---|---|
playback | - |
record | make a recording (quit from the menu before DURATION, so its playback ends there), then copy /var/tmp/armagetronad-bench.aarec into assets/ (the asset of 2026-10-17 is a 90 s one, cut) |
live | the live game for 45 s, no window (the old smoke test) |
texinv | Mesa 0020 off (texture caches invalidated on every draw, as before ps3.262): bisects its cost |
relocs | Mesa ps3.282 switches (docs/archive/GAME-PLAN.md): relocations back on (0019 off), or 0019 kept with a submission every N draws (0026) |
kick150 | - |
kick100 | - |
kick50 | - |
Conditions: what a variant adds to the recipe (its environment switches) and what ran alongside (the stall recorder, a profiler). Rows marked invalid or not comparable say why. The other numbers, the build and the run length are under "All numbers" in each batch.
What the run length means
"warm-up + 120 s" is a steady-state run, which leaves out a 60 s warm-up and measures 120 s or more; "whole run" counted everything from the start, the way runs before 2026-10-17 did.
Suite summary
2026-10-20 · batch suite-summary-2026-10
Purpose: The 13 benchmarked games (15 rows) with their known numbers, the rules-1 calibration of the rating formula (docs/GAMEDB.md "Calibration", tools/gamedb/score.py). Each value names its source: a batch's cell, or the suite table of docs/games/README.md where the number is a reading or a range (no single cell holds it). Stutter is the stall seconds over the 2-minute window; load unmeasured, stability assumed clean, picture and display right.
| Row | ram_free MiB | vram MiB | fps_steady frames a second | stall_play_s seconds | low1_play frames a second | longest_play_ms ms | sound_faults_per_min faults a minute |
Picture, sound |
|---|---|---|---|---|---|---|---|---|
| Armagetron base | 58 | 28 | 83.0 | 0.1 | 24.1 | 127 | 0 | right, right |
Armagetron live AI rounds (recipe e5a556d), Mesa ps3.408; kick
2026-10-18 · batch 2026-10-18-armagetron-live-ai · 1 run in the batch (all games)
| Variant | Conditions | Steady fps in the window |
1% low fps | Stalls frames over 100 ms |
|---|---|---|---|---|
base |
waker kick at 10 ms | 83 | 24.1 | 1 |
All numbers and setup of this batch
| Variant | Numbers | Setup |
|---|---|---|
base |
fps whole run 83 cpu 144.1 seconds underruns 289 underruns ram_free 58.0 MiB vram 27.5 MiB longest_ms 127.1 ms stall_play_s 0.1 seconds longest_play_ms 127.1 ms | Mesa ps3.408 the game's resolution, warm-up + 120 s |
Observations: One run. Live AI rounds spectated (first-start screens skipped, virtual keys): steady 83 fps, 1 stall (127 ms). The webcam showed the arena with the AI light cycles. Not comparable with the old playback numbers.
Mesa ps3.299 + ps3rsx 0.249
2026-10-17 · batch 2026-10-17-mesa-299 · 24 runs in the batch (all games)
| Variant | Conditions | Steady fps in the window |
1% low fps | Stalls frames over 100 ms |
|---|---|---|---|---|
base |
the recipe as it is | - | - | - |
All numbers and setup of this batch
| Variant | Numbers | Setup |
|---|---|---|
base |
fps whole run 61.5 cpu 63.2 seconds underruns 41.5 underruns ram_free 53.1 MiB vram 28.8 MiB | Mesa ps3.299, ps3rsx 0.249 the game's resolution, whole run |
Observations: - Clean runs only, two per game, watched by a webcam script with a single SSH connection. - Armagetron: 61.5 fps (50 on ps3.282). - OpenArena (compressed textures): 52.4 (51.3). - ClassiCube: 46.3, against 46.9 on ps3.282 on the same boot: noise. - X-Moto: 44.5, against 47.0 on ps3.282 on the same boot: a real -5%. Turning off adaptive submission (45.5) or the index change (45.0) does not restore it. Fragment constants from VRAM (46) and the late index upload (46) each win back about a frame: 0031 and 0028 cost about one frame each. - 0031 checked on the webcam with OpenArena's OpenGL 2 renderer: right picture, no GPU hang. The driver tests pass. - Earlier runs that day, made with an old webcam watcher that logged in over SSH every 5 s, were slowed and are left out.
Mesa ps3.282 + ps3rsx 0.249: Armagetron switches (relocs and kick N)
2026-10-17 · batch 2026-10-17-mesa-282-armagetron · 10 runs in the batch (all games)
| Variant | Conditions | Steady fps in the window |
1% low fps | Stalls frames over 100 ms |
|---|---|---|---|---|
base |
the recipe as it is | - | - | - |
kick100 |
NV30_PS3_KICK_DRAWS=100 | - | - | - |
kick150 |
NV30_PS3_KICK_DRAWS=150 | - | - | - |
kick50 |
NV30_PS3_KICK_DRAWS=50 | - | - | - |
relocs |
NV30_PS3_RELOCS=1 | - | - | - |
All numbers and setup of this batch
| Variant | Numbers | Setup |
|---|---|---|
base |
fps whole run 50.0 cpu 59.5 seconds underruns 26.5 underruns ram_free 62.5 MiB vram 28.6 MiB | Mesa ps3.282, ps3rsx 0.249 the game's resolution, whole run |
kick100 |
fps whole run 60.5 cpu 62.8 seconds underruns 31.0 underruns ram_free 52.4 MiB vram 28.8 MiB | Mesa ps3.282, ps3rsx 0.249 the game's resolution, whole run |
kick150 |
fps whole run 58.5 cpu 56.5 seconds underruns 17.5 underruns ram_free 59.0 MiB vram 28.6 MiB | Mesa ps3.282, ps3rsx 0.249 the game's resolution, whole run |
kick50 |
fps whole run 60.5 cpu 63.5 seconds underruns 25.0 underruns ram_free 59.9 MiB vram 28.8 MiB | Mesa ps3.282, ps3rsx 0.249 the game's resolution, whole run |
relocs |
fps whole run 60.5 cpu 62.1 seconds underruns 23.0 underruns ram_free 64.3 MiB vram 28.8 MiB | Mesa ps3.282, ps3rsx 0.249 the game's resolution, whole run |
Observations: - Mesa ps3.282 with a run-time switch per patch. Two interleaved rounds of Armagetron, 90 s each, HUD frame rate. - As built: 50 fps. - Relocations back on (patch 0019 off): 60.5 fps. - A submission every 150 draws (patch 0026): 58.5. Every 100 draws: 60.5. Every 50 draws: 60.5. - CPU time is the same in all variants (57-64 s). - The runs show the effect, not the mechanism. Two runs per variant.
Mesa ps3.246 control for Armagetron (same boot as ps3.262)
2026-10-17 · batch 2026-10-17-mesa-246-control · 2 runs in the batch (all games)
| Variant | Conditions | Steady fps in the window |
1% low fps | Stalls frames over 100 ms |
|---|---|---|---|---|
base |
the recipe as it is | - | - | - |
All numbers and setup of this batch
| Variant | Numbers | Setup |
|---|---|---|
base |
fps whole run 60.0 cpu 63.2 seconds underruns 21.5 underruns ram_free 61.6 MiB vram 28.6 MiB | Mesa ps3.246 the game's resolution, whole run |
Observations: - Mesa ps3.246 reinstalled on the same boot as the ps3.262 runs. - Armagetron: 61 and 59 fps (HUD median), against 51-53 on ps3.262 in the previous batch. - The second run also printed the game's own playback average, 71.2 fps. That is a different measure and is not used. - This confirms the loss comes from Mesa ps3.262's patches, not from the boot or the day.
Mesa ps3.262 repeats and Armagetron without 0020 (texinv)
2026-10-17 · batch 2026-10-17-mesa-262-repeats · 8 runs in the batch (all games)
| Variant | Conditions | Steady fps in the window |
1% low fps | Stalls frames over 100 ms |
|---|---|---|---|---|
base |
the recipe as it is | - | - | - |
texinv |
NV30_TEX_INVALIDATE_ALWAYS=1 | - | - | - |
All numbers and setup of this batch
| Variant | Numbers | Setup |
|---|---|---|
base |
fps whole run 53.0 cpu 60.8 seconds underruns 29.0 underruns ram_free 67.8 MiB vram 28.3 MiB | Mesa ps3.262 the game's resolution, whole run |
texinv |
fps whole run 52.0 cpu 61.1 seconds underruns 23.5 underruns ram_free 62.7 MiB vram 28.8 MiB | Mesa ps3.262 the game's resolution, whole run |
Observations: - Same boot as the Mesa ps3.262 suite round. - QuakeSpasm again: 73.0 and 84.1 fps (66.3 in the suite round). Single QuakeSpasm runs spread widely on the same Mesa. - Neverball again: 77.7 and 74.2 fps (71.5 in the suite round). - Armagetron with ps3.262 as built: 53 and 53 fps. With patch 0020 undone (the texture cache flushed before every draw again): 52 and 52. So the drop from 60 to 52 is not patch 0020. - All Armagetron numbers here are the Gallium HUD's median. Two runs per variant.
Mesa ps3.262 + ps3rsx 0.249
2026-10-17 · batch 2026-10-17-mesa-262 · 12 runs in the batch (all games)
| Variant | Conditions | Steady fps in the window |
1% low fps | Stalls frames over 100 ms |
|---|---|---|---|---|
base |
the recipe as it is | - | - | - |
All numbers and setup of this batch
| Variant | Numbers | Setup |
|---|---|---|
base |
fps whole run 51 cpu 59.3 seconds underruns 23 underruns ram_free 71.7 MiB vram 28.8 MiB | Mesa ps3.262, ps3rsx 0.249 the game's resolution, whole run |
Observations: - The patches: no relocations, the texture cache invalidated only after texture changes, no duplicate vertex-format writes, batched vertex program constants. One run per game. - Against ps3.246: OpenArena with compressed textures 44.2 to 54.4 fps (+23%), Yamagi 51.4 to 63.9 (+24%), X-Moto 39 to 47 (+21%), Extreme Tux Racer 46 to 53 (+15%), dsda-doom software 33 to 38 with the demo now finishing (+15%), ClassiCube 47.6 to 50.1 (+5%). - Neverball 78.2 to 71.5 (noise), QuakeSpasm 70.5 to 66.3 (repeated in the next batch), Luanti 53 to 53. - Armagetron dropped from 60 to 51 (driver frame counter). - BananaBread reads 28 to 8 on the counter, but it was memory-stalled (13 MB free, killed at the cap): not a driver number, and the cause is uncertain. - The kernel's command validator rejected nothing. - Single runs.
Mesa ps3.246 + ps3rsx 0.249
2026-10-17 · batch 2026-10-17-mesa-246-ps3rsx-249 · 12 runs in the batch (all games)
| Variant | Conditions | Steady fps in the window |
1% low fps | Stalls frames over 100 ms |
|---|---|---|---|---|
base |
the recipe as it is | - | - | - |
All numbers and setup of this batch
| Variant | Numbers | Setup |
|---|---|---|
base |
fps whole run 60 cpu 65.0 seconds underruns 52 underruns ram_free 72.4 MiB vram 28.3 MiB | Mesa ps3.246, ps3rsx 0.249 the game's resolution, whole run |
Observations: - One run per game. Frame rates by the game's own counter, or the driver's frame counter where marked: - Armagetron 60 (counter), ClassiCube 47.6, dsda-doom software 33 (counter, demo unfinished), Extreme Tux Racer 46 (counter), Luanti 53 (counter), Neverball 78.2. - OpenArena with compressed textures 44.2 (longest frame 2.1 s), QuakeSpasm 70.5, BananaBread with CPU skinning 28 (counter), X-Moto 39 (counter), Yamagi Quake II 51.4. - Pictures right. Sound right except dsda-doom (1161 underruns, MIDI music), BananaBread (176) and OpenTTD (1429, choppy). - OpenTTD used 55.5 s of CPU and has no frame rate (plain X11 path). - Single runs: differences of a few fps are within the run-to-run spread.
Group 1 on Mesa ps3.225: ClassiCube demo and Armagetron playback
2026-10-16 · batch 2026-10-16-group1-mesa-225 · 2 runs in the batch (all games)
| Variant | Conditions | Steady fps in the window |
1% low fps | Stalls frames over 100 ms |
|---|---|---|---|---|
base |
the recipe as it is | - | - | - |
All numbers and setup of this batch
| Variant | Numbers | Setup |
|---|---|---|
base |
fps whole run 64.0 cpu 66.0 seconds underruns 6 underruns ram_free 78.2 MiB vram 28.8 MiB | Mesa ps3.225 the game's resolution, whole run |
Observations: - ClassiCube fly-through (3000 fixed frames): 43.24 fps by its own counter, longest frame 278 ms, load 12.65 s, 78 s of CPU. - Armagetron playback of the 4-AI round completed for the first time: 63.96 fps by the game's own average, ended normally at 94 s. - The game's own average differs from the driver frame counter median used in later Armagetron batches (50 to 61). - Single runs.
Armagetron benchmark recording (Mesa ps3.223)
2026-10-16 · batch 2026-10-16-armagetron-recording · 2 runs in the batch (all games)
| Variant | Conditions | Steady fps in the window |
1% low fps | Stalls frames over 100 ms |
|---|---|---|---|---|
base |
the recipe as it is | - | - | - |
record |
the recipe as it is | - | - | - |
All numbers and setup of this batch
| Variant | Numbers | Setup |
|---|---|---|
base |
fps whole run 48 underruns 101 underruns ram_free 71.9 MiB vram 28.8 MiB | Mesa ps3.223 the game's resolution, whole run |
record |
fps whole run 50 cpu 72.1 seconds underruns 78 underruns ram_free 64.9 MiB vram 28.3 MiB | Mesa ps3.223 the game's resolution, whole run |
Observations: - The recording run spectated a 90 s round of 4 AI players: 50 fps on the driver's frame counter, 78 underruns. - A first playback launch was killed without a result (48 fps on the counter). - Playback first completed on ps3.225 (the following batches).
Mesa ps3.223 + ps3rsx 0.219
2026-10-16 · batch 2026-10-16-mesa-223 · 5 runs in the batch (all games)
Not comparable: every game was profiled with perf during the run, which slows it.
| Variant | Conditions | Steady fps in the window |
1% low fps | Stalls frames over 100 ms |
|---|---|---|---|---|
basenot comparable |
the recipe as it is | - | - | - |
All numbers and setup of this batch
| Variant | Numbers | Setup |
|---|---|---|
base |
fps whole run 102 cpu 21.8 seconds underruns 22 underruns ram_free 33.2 MiB vram 26.5 MiB | Mesa ps3.223, ps3rsx 0.219 the game's resolution, whole run |
Observations: - Each game was profiled for 15 s, so the fps are lower than in plain runs (the run records don't show the profiling). One run per game. - X-Moto 34.6 fps (own counter), 208 underruns. - OpenArena with compressed textures 37.2 fps, longest frame 3.1 s. - ClassiCube 54, Extreme Tux Racer 40 (in its menu), Armagetron 102 (a live first-run game); all three on the driver's frame counter. - Result: Mesa took 28 to 56% of the CPU and ps3rsx 6 to 16%.
Group 1 smokes ETR and Armagetron on Mesa ps3.223
2026-10-16 · batch 2026-10-16-etr-armagetron-smokes · 3 runs in the batch (all games)
| Variant | Conditions | Steady fps in the window |
1% low fps | Stalls frames over 100 ms |
|---|---|---|---|---|
base |
the recipe as it is | - | - | - |
All numbers and setup of this batch
| Variant | Numbers | Setup |
|---|---|---|
base |
fps whole run 89 cpu 13.8 seconds underruns 35 underruns ram_free 83.6 MiB vram 27.7 MiB | Mesa ps3.223 the game's resolution, whole run |
Observations: - Extreme Tux Racer ran to the 45 s cap: 42 fps on the driver's frame counter, picture right, no kernel errors. - Armagetron's first launch quit after 5 s on "Unknown command line option --fullscreen", an error in our recipe. - The second launch ran to the 45 s cap at 89 fps on the first-run screen, which is not a benchmark. - Single runs; the numbers are driver counter medians.