BENCHMARK(5)

📦 version 7.44.0, in nixpkgs unstable

🎀 benchmark(5) — the BRL-CAD Benchmark Suite Methodology, bestie edition 🎀

okay so this is the manual page for benchmark, aka the BRL-CAD Benchmark Suite Methodology, and it also explains the Ray-Tracing Figure of Merit (RTFM) — which is the whole vibe-check system used to judge how fast different computer systems actually are. no cap, it's about to be a whole thing 💅

✨ INTRODUCTION ✨

so the core of this entire test is the rt ray-tracing program. that's the main character here, periodt. 🫶

the results for each processor are figured out by running a benchmark test of the RT ray-tracing program on each of the sample databases provided. and listen — rt is kinda picky about its inputs and outputs:

  • it needs a binary database for input (db/xxx.g)
  • it produces a binary image as output (bench/xxx.pix)
  • and it also drops a log file (bench/xxx.log)

every single run produces a 512x512x24 bit color shaded image in a bench/xxx.pix file, plus a run log in bench/xxx.log. reference copies of all of this are provided as pix/xxx.pix and pix/xxx.log, so you can compare your results against the gold standard. ⭐

the sample databases are processed in order of increasing difficulty (they said what they said, it gets harder as you go 😭), and each one produces a colorful image. in every picture there are two light sources:

  • the primary source is located near the center of the model (the white ball)
  • the secondary light source is located at the "eye" position

🌈 THE BENCHMARK MODELS 🌈

here's the lineup, and honestly? it's giving museum collection:

  • moss.g — a simple model containing a yellow torus, a green ellipsoid, a bluish cube, and a pink truncated (distorted) cone, all sitting on a blue plate. take note of the shadowing and the "specular splash" — it's iconic, you'll know it when you see it ✨
  • world.g — same model as moss.g, but with various rendering features enabled. the model is enclosed in a hollow cloud sphere. the plate is a mirror, the egg is crystal, and the box has a debugging texture map applied. fancy behavior fr fr 💅
  • star.g — a familiar sight to Star Trek fans: a low-detail exterior of the ship "Enterprise". the hull is white, with some other structures in other colors. the small red object in the foreground is the shuttlecraft Galileo, shown to scale 🖖
  • bldg391.g — an exploded view of a two-story imaginary computer site, including various walls, red stairways, green disk drives, etc. the lower floor is a mirror, and the upper floor is glass. very architectural, very demure 🏢
  • m35.g — a truck. that's it. that's the description. it's a truck. 🚚
  • sphflake.g — a sphereflake. (yes, that's the whole description, and honestly? obsessed. ❄️)

🦋 MACHINE INDEPENDENCE 🦋

the databases are provided in an ASCII format in the db directory, and the ascii-to-binary converters live in the conv directory. on most systems, the command make benchmark (see the document Installing the BRL-CAD Package) will automatically compile everything, generate the binary .g files, and run the benchmark tests. one command, whole pipeline, slay. 🫶

the reference images are distributed in binary BRL-CAD pix(5) format. earlier releases distributed the reference images in the portable ASCII form and then converted them to the binary form — but that consumed a significant amount of extra storage on UNIX machines, so this practice was discontinued. so long, extra storage 😭

it is assumed that anyone attempting to benchmark a system that can not read pure binary files will at least have access to a UNIX system. that UNIX system can be used to convert the reference images to the portable ASCII form with the pix2asc(1) tool, and then those images may be sent to the machine under test. teamwork makes the dream work 💖

ngl, if you encounter difficulty with running the benchmark, you might wish to build the full BRL-CAD package — that way you can take advantage of the image tools in determining the nature of the problem. for example, one system suffered from a compiler bug that made TGC solids vanish, while another system had difficulties with the light visibility calculations. real-world chaos, bestie 💀

🍓 OBTAINING CORRECT RESULTS 🍓

the benchmark timings are not considered valid unless the correct results are given. like, the numbers don't count if the picture's wrong, periodt. 🙈

make sure the answers match the reference files to within plus-or-minus one in each color (see pixdiff(1)). slight variations in the calculated pixels are generally attributable to variations in floating point precision — totally normal, lowkey expected.

you may also wish to compare the execution times and the log file remarks from your tests (in the bench/xxx.log files) with the VAX-11/780 (with hardware FPA) times, which are given in the pix/xxx.log files. heads up though: the log file for sphflake has yet to be run on the same system, causing weighted results. it's still useful and informative, however, for deriving performance statistics. (iconic of it to still be helpful tbh 👑)

the reference images are also located in the pix/ directory, in case you were wondering. ✨

for example, if you are running the benchmark on a Cray XMP, the world.pix and bldg391.pix images will have a single byte in the blue channel off by 1, out in the cloud background. this is not to be considered an error — do NOT come for it, it's fine 💅

🌸 BENCHMARK IMAGE DISPLAY 🌸

if you have a suitable framebuffer, you may wish to display the images generated by the benchmark and compare them to the reference images. cute, visual, satisfying. 🎀

this section assumes that you have successfully compiled the full set of BRL-CAD software, not just the benchmark portion. programs for dealing with images are in the util directory.

to display a pix(5) file on a framebuffer, set the environment variable FB_FILE (see brlcad(1) for details) and run pix-fb. note that by leaving FB_FILE unset, your default display will be used. easy.

cd bench
pix-fb moss.pix

if the images computed on your machine do not match the reference images, the program util/pixdiff will compute a pix file that highlights the differences for you, and report a summary of bytes equal, off-by-1, and off-by-many. e.g.

pixdiff moss.pix ../pix/moss.pix (or) pix-fb

for a display of the relative magnitude of the differences at each pixel, use util/bwdiff instead, e.g.

bwdiff moss.pix ../pix/moss.pix (or) pix-fb

or

bwdiff moss.pix ../pix/moss.pix (or)
bwmod -s 128 -m 4 -a 128 | pix-fb

and for statistics on the differences, use

bwdiff moss.pix ../pix/moss.pix (or) pixstat

💅 THE COMPETITION 💅

okay so real talk bestie: the full benchmark takes about 9 CPU hours on a VAX-11/780, so like… scale your expectations accordingly, mkay? this is a marathon not a sprint, no cap.

here's a lil sampling of some systems that have been tested, with as much configuration info as could be obtained. notice how some systems have a great many options for small things that can affect performance, like memory interleaving — yeah, those are the messy ones 🙈.

  • _ — _
  • axposf — DEC Alpha 4000/710, 256M, OSF/1 V3.0(1)
  • ex-art — Pentium-P90, 16MB, 256K cache, BSD/OS 2.0,
  • hawks — 486DX/33, 24MB, 256K cache, BSD/OS 2.0
  • orac — 486/100, 20MB, No cache, BSD/OS 2.0 (laptop)
  • pk — Cray C90, C916/161024, UNICOS 8.0.2.4
  • vapor — SGI, 18* 75 MHz R8000 IP21, 1024M RAM 8-way, Irix 6.0
  • vblt — 66 MHz 486DX2/66, 16M RAM, Linux 1.1.73
  • vgr — DEC VAX-11/780 w/FPA, 64M RAM, 4.3 BSD
  • vision — Sun-3/280 w/68881 FPA, 16M RAM, SunOS 4.1.1
  • voyage — SGI, 4* 25 MHz R3000 IP7, 56M RAM, Irix 4.0.1
  • whiz — SGI Ingido**2, 100 MHz IP22, 96M RAM, Irix 5.2
  • waffle — SGI, 4* 25 MHz R3000 IP7, 64M RAM, Irix 4.0.5 (heavy network use)
  • warp — Sun SPARCstation 5, 32M RAM, SunOS 4.1.4
  • wax — SGI, 24* 150 MHz R4400 IP19, 1024M RAM 4-way, Irix 5.2
  • wilson — SGI, 8* 100 MHz R4400 IP19, 1024M RAM 8-way, Irix 5.2
  • wilted — Sun-4/25 ELC, 16M RAM, SunOS 4.1.1
  • wimp — 33 MHz 386DX/33 w/387, 256k cache, 16M RAM, BSDI 2.0beta2
  • wimpy — Sun SS10BSX-GX, 2* 50 MHz SuperSPARC, 64M RAM, SunOS 5.4
  • wizard — SGI Ingido**2, 150 MHz IP22, 96M RAM, Irix 5.3

(1) the Alpha was heavily loaded while the benchmark was being run 😭 and DEC literally requires us to show the load average at the time, so here it is in all its chaotic glory:

03:28 up 18 days, 10:59, 7 users, load average: 2.00, 2.00, 2.08

🔥 BENCHMARK RESULTS 🔥

in the tradition of Dongarra, the Rays/sec figure is considered the "RT Figure of Merit" (RTFM) ✨ iconic naming, we stan.

⚠️ bestie please internalize this: the RTFM can only be compared between different runs on the same database. it is not meaningful to compare RTFM numbers between different databases. like, don't do it. fr fr.

the Rays/sec number for a multi-processor machine is for "aggregate CPU cluster seconds", rather than rays/total CPU seconds (which remains fairly constant on good parallel machines). slay, parallel machines 🫶

the statistics are recorded in an easily understood format, with all results for a particular configuration listed on a single line — we love a tidy queen. the numbers reported are the rays/sec (RTFM) figure.

  • entries denoted with an asterisk ("*") produced incorrect results 💀
  • entries denoted with a dash ("-") were not available and should be considered differently than those entries with an asterisk 🙈

generally the results are not available because time constraints prevented running the complete set of databases. literally just a scheduling issue, not a vibe issue. periodt.

🌈 RELEASE 4.4 STATISTICS (4-JAN-1995) 🌈

Okay bestie, so here's the tea on Release 4.4: there is very little difference in RTFM numbers between Release 4.0 and Release 4.4, which honestly means they can be usefully compared. Slay. 🎀

But — and this is the part where we hold hands — in some cases the 4.4 performance is slower due to the use of shared libraries. Like, for example, "vision" dropped from 1.33 to 1.22 due to the shared libraries. 😭 For most users, reducing the disk consumption from 300 MBytes to 30 MBytes makes this a reasonable tradeoff. However, for the best performance, use non-shared libraries. Noted. 💅

The numbers below are straight from the original, in the exact order they appear. The columns are H/W, Moss, World, Star, Bldg, M35, Mean, and S/W Notes — and yes, one row is literally just a .TH sitting alone in the H/W column with nothing else filled in. That's exactly what's in the original. We don't make the rules, we just report them. 🦋

  • .TH — Moss: (blank) · World: (blank) · Star: (blank) · Bldg: (blank) · M35: (blank) · Mean: (blank) · S/W notes: (blank)
  • vapor — Moss 810.91 · World 1091.39 · Star 1012.63 · Bldg 1027.25 · M35 1092.85 · Mean 1007 · S/W notes: -P18 static
  • vapor — Moss 773.8 · World 1038.7 · Star 979.49 · Bldg 970.91 · M35 1036.62 · Mean 959.9 · S/W notes: -P17
  • vapor — Moss 732.85 · World 1021.69 · Star 968.93 · Bldg 922.62 · M35 1024.15 · Mean 934.04 · S/W notes: -P18 (SW lock)
  • wax — Moss 769.38 · World 940.83 · Star 953.36 · Bldg 924.7 · M35 1050.14 · Mean 927.68 · S/W notes: -P24
  • vapor — Moss 730.17 · World 994.97 · Star 934.43 · Bldg 917.1 · M35 987.83 · Mean 912.9 · S/W notes: -P16
  • wax — Moss 725.26 · World 957.68 · Star 927.61 · Bldg 932.67 · M35 992.75 · Mean 907.19 · S/W notes: -P24
  • wax — Moss 756.95 · World 915.96 · Star 940.43 · Bldg 903.45 · M35 966.09 · Mean 896.57 · S/W notes: -P23
  • vapor — Moss 686.8 · World 939.63 · Star 897.72 · Bldg 884.39 · M35 927.16 · Mean 867.14 · S/W notes: -P15
  • wax — Moss 691.18 · World 853.75 · Star 887.17 · Bldg 859.32 · M35 933.41 · Mean 844.96 · S/W notes: -P21
  • wax — Moss 724.73 · World 683.08 · Star 890.87 · Bldg 900.82 · M35 939.31 · Mean 827.76 · S/W notes: -P22
  • vapor — Moss 645.32 · World 881.18 · Star 843.23 · Bldg 801.41 · M35 859.26 · Mean 806.08 · S/W notes: -P14
  • wax — Moss 637.28 · World 757.74 · Star 820.21 · Bldg 871.37 · M35 877.6 · Mean 792.84 · S/W notes: -P20
  • vapor — Moss 608.12 · World 835.84 · Star 795.36 · Bldg 759.95 · M35 806.32 · Mean 761.11 · S/W notes: -P13
  • wax — Moss 606.98 · World 784.3 · Star 754.46 · Bldg 799.95 · M35 794.52 · Mean 748.04 · S/W notes: -P19
  • wax — Moss 590.16 · World 758.23 · Star 794.06 · Bldg 794.52 · M35 757.79 · Mean 738.95 · S/W notes: -P18
  • wax — Moss 579.29 · World 722.35 · Star 787.55 · Bldg 720.26 · M35 754.65 · Mean 712.82 · S/W notes: -P17
  • vapor — Moss 571.14 · World 783.72 · Star 740.47 · Bldg 697.24 · M35 746.72 · Mean 707.85 · S/W notes: -P12
  • wax — Moss 553.72 · World 697.1 · Star 744.21 · Bldg 663.61 · M35 723.03 · Mean 676.33 · S/W notes: -P16
  • vapor — Moss 534.85 · World 716.58 · Star 685.87 · Bldg 656.25 · M35 685.77 · Mean 655.86 · S/W notes: -P11
  • wax — Moss 519.01 · World 655.57 · Star 690.95 · Bldg 651.9 · M35 663.92 · Mean 636.27 · S/W notes: -P15
  • wax — Moss 475.88 · World 637.33 · Star 665.38 · Bldg 608.31 · M35 653.99 · Mean 608.17 · S/W notes: -P14
  • vapor — Moss 499.46 · World 663.26 · Star 633.48 · Bldg 614.58 · M35 624.34 · Mean 607.02 · S/W notes: -P10
Heads up bestie: those .TH and blank cells above aren't me being lazy, that's genuinely what the original table has there. Then below the fold, the row where S/W Notes reads -P18 / static is split across two lines in the source — I've got the whole thing on one line for you so it's easier to read. 🫶

The vibes? vapor takes the top spots, wax hangs around the middle, and the whole thing ends on a vapor row with a Mean of 607.02. Iconic. 👑

okay bestie, buckle up 💅 this is the part where the page turns into The Wall Of Numbers — and honestly? it ate. it's giving "receipts" energy. let's get into it.

the table is running big → small (like, the rows go from the chonkiest times at the top down to the babies at the bottom), and each row is: model name, then six numbers, then the -P flag for that run. no cap, that's the whole shape of it. keeping every digit exactly as it stands, because we do not round in this house 🙈

🧚 THE ROWS, IN ORDER 🧚

  • wax — 449.16 · 591.54 · 616.21 · 590.59 · 614.25 · 572.35 — -P13
  • vapor — 463.3 · 598.34 · 575.07 · 555.62 · 560.89 · 550.64 — -P9
  • wax — 424.71 · 556.8 · 595.02 · 521.63 · 538.78 · 527.38 — -P12
  • vapor — 424.89 · 534.5 · 509.85 · 497.63 · 503.99 · 494.17 — -P8
  • wax — 391.66 · 490.63 · 545.28 · 488.57 · 508.96 · 485.02 — -P11
  • wax — 315.45 · 472.03 · 525.47 · 442.37 · 461.44 · 443.35 — -P10
  • vapor — 381.1 · 467.67 · 440.13 · 425.57 · 442.04 · 431.3 — -P7
  • wax — 309.35 · 426.37 · 386.11 · 417.85 · 434.96 · 394.92 — -P9
  • vapor — 336.55 · 392.2 · 390.72 · 373.92 · 380.03 · 374.68 — -P6
  • wax — 307.6 · 361.32 · 396.07 · 380.17 · 380.74 · 365.18 — -P8
  • wax — 245.87 · 332.33 · 362.5 · 338.62 · 331.39 · 322.14 — -P7
  • vapor — 288.45 · 327.78 · 328.76 · 314.54 · 320.16 · 315.93 — -P5
  • wax — 254.82 · 255.29 · 299.93 · 302.08 · 256.54 · 273.73 — -P6
  • wilson — 223.01 · 265.17 · 287.8 · 262.19 · 269.62 · 261.55 — -P8
  • vapor — 240.6 · 269.93 · 266.46 · 253.99 · 259.05 · 258 — -P4
  • wax — 215.96 · 231.65 · 221.78 · 248.63 · 221.45 · 227.89 — -P5
  • wax — 177.94 · 185.27 · 213.74 · 201.64 · 197.27 · 195.17 — -P4
  • vapor — 187.58 · 198.84 · 202.28 · 191.94 · 194.74 · 195.07 — -P3
  • wax — 120.4 · 138.01 · 162.12 · 159.36 · 141.33 · 144.24 — -P3
  • vapor — 130.61 · 140.04 · 134.88 · 129.38 · 129.78 · 132.93 — -P2
  • wax — 100.83 · 102.09 · 111.12 · 87.26 · 98.62 · 99.98 — -P2
  • wimpy — 99.10 · 95.78 · 92.15 · 97.51 · 96.65 · 96.23 — -P2
  • wimpy — 82.14 · 93.60 · 92.17 · 103.59 · 82.22 · 90.74 — -P2
  • axposf — 81.05 · 82.27 · 86.72 · 84.36 · 79.28 · 82.73 — -P1
  • vapor — 79.68 · 77.7 · 73.89 · 71.38 · 70.06 · 74.54 — -P1 static

a few things worth pointing at with the whole hand, because the details are chef's kiss:

  • the top of the list is straight-up wax and vapor trading places at -P13, -P9, -P12, -P8, -P11, -P10, -P7… like, they are FIGHTING for the podium 💀
  • our one and only wilson cameo sits right there at -P8 with 223.01 · 265.17 · 287.8 · 262.19 · 269.62 · 261.55 — icon behavior, no notes 👑
  • wimpy shows up twice, both at -P2, and both times the numbers are hovering in the 82–103 zone
  • axposf clocks in once at -P1 with 81.05 · 82.27 · 86.72 · 84.36 · 79.28 · 82.73
  • and the very last row — vapor at -P1 — is the only one rocking that extra static note on the side. not me zooming in on it 👀 lowkey the main character of the bottom of this table

also, gentle reminder since we're deep in number soup: the whole point of the machine independence section earlier is that these digits are only worth comparing the way the page says to compare them — so don't go rogue and compare things out of context, fr fr 🫶

okay that's this chunk — the data dump continues and there's more coming, so keep scrolling bestie 🌸

...okay bestie we are STILL in the results table, no we're not done, this thing is long and honestly? iconic for it. 💅

Same column order as the top of the table — nothing moved, nothing re-sorted, I'm just reading the board from the jump.

  • vapor — 69.34, 70.14, 67.52, 65.55, 66.69, 67.84 — flag column says -P1 and shared
  • wizard — 63.8, 55.86, 68.97, 73.21, 61.56, 64.68 — flag column empty
  • wax — 62.28, 58.58, 65.24, 67.39, 59.31, 62.56 — -P1
  • wax — 59.08, 56.06, 61.45, 64.4, 56.11, 59.42 — -P1
  • voyage — 52.78, 55.44, 58.19, 58.03, 52.41, 55.37 — -P4
  • pk — 64.89, 55.71, 48.01, 49.50, 53.43, 54.30 — -P1
  • waffle — 38.56, 47.63, 62.36, 59.62, 51.20, 51.87 — -P4
  • wimpy — 50.17, 48.82, 50.46, 55.12, 49.27, 50.76 — -P1
  • wimpy — 47.80, 43.23, 51.37, 54.10, 46.23, 48.54 — -P1
  • weber — 46.53, 43.45, 50.88, 50.33, 46.03, 47.44 — -P1
  • wilson — 41.16, 39.54, 45.98, 42.93, 34.82, 40.88 — -P1
  • whiz — 36.25, 31.08, 37.21, 39.62, 36.98, 36.22 — flag column empty
  • warp — 19.88, 21.33, 23.71, 24.19, 22.60, 22.34 — flag column empty
  • ex-art — 16.89, 18.54, 22.98, 26.97, 25.57, 22.19 — flag column empty
  • orac — 7.57, 8.23, 10.72, 12.43, 11.30, 10.05 — flag column empty
  • wilted — 9.58, 9.53, 10.66, 10.53, 9.87, 10.03 — flag column empty
  • hawks — 3.64, 4.17, 5.71, 6.90, 6.23, 5.33 — flag column empty
  • vblt — 4.85, 4.89, 5.29, 5.24, 4.90, 5.03 — flag column empty
  • wimp — 1.16, 1.34, 1.82, 2.25, 2.06, 1.72 — flag column empty
  • vision — 1.07, 1.23, 1.23, 1.31, 1.26, 1.22 — -P1
  • vgr — 1.0, 1.0, 1.0, 1.0, 1.0, 1.0 — flag column empty

A few little things I'm obsessed with here, and you should be too:

  • vapor is sitting at the top of this chunk AND it's the only row that has two things going on in that last column — -P1 on one line, shared on the next. Miss thing is doing the most ✨
  • wax shows up twice and so does wimpy — two rows each, both flagged -P1. Don't merge them in your head, they're separate lines with separate numbers, and yes I typed them exactly as they are 😤
  • voyage and waffle are your -P4 girlies in this stretch — nobody else down here is -P4
  • wizard has 63.8, not 63.80. wax has 64.4. vgr has 1.0 six times over. do not "helpfully" add zeros, I am begging you 🙏
  • vgr literally ends the whole board with 1.0, 1.0, 1.0, 1.0, 1.0, 1.0 — every column, same value. no cap. make of that what you will, I'm just the messenger 🫶

The board finally bottoms out at vgr, and that's not the end of the page, bestie — there's more coming, don't you dare close the terminal 👀

💅 THE OLD COMPETITION

Okay bestie, so these are the definitions for the systems that got measured in previous tests — like, the vintage lineup, the OGs. 🕰️ No cap, this is basically a museum roster, and we're keeping every single spec exactly as it was recorded. Don't come for me about the units or the spellings; I'm copying them character for character because that's the whole point. ✨

Heads up: the original page lays all of this out in a big table with the columns System Name and Configuration. Tables aren't the vibe here, so I'm giving you the exact same info as a list — same names, same configs, nothing dropped. 🫶

The table also has a blank row with just _ and _ in it — that's literally in the original, so here it is, preserved for the culture:

  • _ — _

And now, the lineup 👑

  • alliant — Alliant FX/8 (8 CEs, 9 IPs, 64 Mbytes), Concentrix 3.0
  • amber — HP 9000/720, 32 MB memory
  • amsaa-seer — Gould PN 9080, w/MACCs, UTX 2.0, 4x4Mb mem boards
  • ardec-3 — Pyramid 90Mx, Dual-CPU, OSx 2.5, 16 Mbytes
  • bob — Cray-2, SN 2009, 4.3ns clock
  • cor3 — Pyramid MIS-12T/3, OSx 192
  • patton — Cray X-MP/48, SN213, UNICOS 5.0, 8.5ns clock
  • uy1 — Cray Y-MP8/2128, UNICOS 5.1.11
  • shpcrc2 — IBM RS/6000, AIX 3.1
  • slc1 — Macintosh II, 68020, 68881, AUX 1.1, GCC 1.39
  • spark — Gould PN 9050, no MACC, UTX 2.0, 2x4Mb mem boards
  • sws2 — Convex C120, 16 Mbytes, Convex Unix 6.2.32.2
  • vector — Alliant FX/80 (8 ACEs, 6 IPs, 32 Mbytes), Concentrix
  • venom — Alliant FX/8 (8 CEs, 6 IPs, 32 Mbytes), Concentrix
  • vgr — VAX 780, FPA, 4.3 BSD
  • virus — Sun-3/50, 15 Mhz clock, 12 Mhz 68881, Sun UNIX 3.2
  • vista — SGI 3030, w/FP chip, UNIX release 3.5
  • vmb — Gould PN 9080, no MACCs, UTX 2.0, 12x1Mb mem boards
  • vhs — Silicon Graphics 4D/60T, 12.5 Mhz clock 16 MB memory
  • ovoyage — Silicon Graphics 4D/120, 16.7 Mhz clock 16 MB memory
  • taylor — Silicon Graphics 4D/220, 25 Mhz clock 16 MB memory
  • crim — Silicon Graphics 4DCRIMS 50Mhz clock 128 MB memory
  • wiltse — Silicon Graphics 4D R4000 50Mhz clock
  • c1east — Convex C1 XP (2 IOPs, 4 Multibus, 64 Mbytes), S/W 6.0.1.12
  • elxsi-gnu — Elxsi 6420, BSD 4.2 16 MB
  • elxsi-gnuy — Elxsi 6420, Sys Vr2, 16MB
  • elxsi-m1 — Elxsi 6410, BSD 4.2 16 MB
  • hep — Denelcor HEP, 4 PEMs
  • indigo — SGI Iris 3030, FPA, GL2-W3.5
  • mseries — MIPS Mseries, Release 0.6, UMIPS 2.1, 16 MB of R2350 RAM. R2000 CPU rev 5.0, R2010 FP rev 2.0
  • multiflow — Multiflow Trace 7/200 (PRELIMINARY)
  • nova — Sun SPARCserver 490, 32 Mb, 33 MHz clock, 33 MHz TI 8847 FPU, SunOS 4.1.1B
  • rh2 — Tektronix, Motorola 88000
  • snm2 — Cray 1-M, SN2, UNICOS 2.0
  • sws1 — Convex C1, ConvexOS 9.1 48MB memory
  • tek4132 — Tektronix 4132, 32082 fpp, UTek 2.3

Lowkey iconic lineup, honestly. The amount of silicon in this list could flatten a building. 💀 But anyway — every name and every config above is exactly as the original page has it, so you can trust it for real. 🌸

  • utah-gr — VAX 785, FPA, 4.3 BSD
  • utah-cs — VAX 8600, FPA, 4.3 BSD
  • veto — Sun 4/260 32MB memory Sun UNIX 4.0.3
  • violet — VaxStationII GPX, Ultrix 1.2
  • vision — Sun-3/280, 16 MB memory
  • vortac — Sun-3/160, 16.67 Mhz clock, 12 Mhz 68881, Sun 3.2
  • walrus — Silicon Graphics 4D/280 25 Mhz clock 64 MB memory Irix 3.3.1
  • whiz — Silicon Graphics 4D/240 25 Mhz clock 56 MB memory Irix 3.3.1
  • whisper — Sun SPARCStation 1+
  • wilted — Sun SPARCstation ELC , Diskless, 12MB memory Sun UNIX 4.1.1
  • worm — Silicon Graphics 4D/280 25 Mhz clock 64 MB memory Irix 3.3.1

💾 OLD BENCHMARK RESULTS 💾

okay so these are the results of previous benchmarks. like, the receipts from back in the day 💅 — the whole point is you can look at what the old machines pulled off and compare, iykyk 🫶

still with me? bestie we are SO not done, keep scrolling 🌸

🌈 RELEASE 1.20 STATISTICS (12-FEB-1987) 🌈

okay bestie, this is the vintage one — the 1987 leaderboard, and it's giving ancient-document energy in the best way possible 💖. Same column setup as the later ones, so here's the map before you scroll:

  • A/* — what flavour of test it was: Rays (the ray-tracing test) or *vgr (the vector-graphics test). The little * up front is just part of how it's written in the table, iykyk.
  • H/W — which machine ran it: tek4132, violet, vgr, indigo, vista, virus, vortac, utah-gr, ardec-3, elxsim1, spark.
  • Moss, World, Star, Bldg — the four benchmark models, and their times.
  • Mean — the average, mostly ? here because people did not come through with the mean, bestie 😭.
  • S/W Notes — the OS and version, plus any little build notes like w/fast_sqrt.

and here's the whole thing, exactly as it appears in the page (it's a lot, so it lives in a code block — don't panic, it's just rows):

┌──────┬─────────┬─────────┬────────┬────────┬────────┬──────┬───────────────┐
│      │         │         │        │        │        │      │               │
│ A/*  │ H/W     │ Moss    │ World  │ Star   │ Bldg   │ Mean │ S/W Notes     │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ .TH  │         │         │        │        │        │      │               │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ Rays │ tek4132 │ 80.5    │ 44.0   │ 33.9   │ 31.4   │ ?    │ UTek 2.3      │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ *vgr │ tek4132 │ 0.72    │ 0.69   │ 0.62   │ 0.62   │ ?    │               │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ Rays │ violet  │ 107.7   │ 57.3   │ 44.7   │ 44.1   │ ?    │ Ultrix        │
│      │         │         │        │        │        │      │ 1.2           │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ *vgr │ violet  │ 0.96    │ 0.90   │ 0.81   │ 0.87   │ ?    │               │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ Rays │ violet  │ 119.0   │ 63.0   │ 50.0   │ 46.3   │ ?    │ Ultrix        │
│      │         │         │        │        │        │      │ 1.2,          │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ *vgr │ violet  │ 1.06    │ 0.99   │ 0.91   │ 0.91   │ ?    │ w/fast_sqrt   │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ Rays │ vgr     │ 112.1   │ 63.6   │ 55.0   │ 50.7   │ ?    │ BSD 4.3       │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ *vgr │ vgr     │ 1.0     │ 1.0    │ 1.0    │ 1.0    │ 1.0  │               │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ Rays │ indigo  │ 115.9   │ 85.3   │ *      │ 101.8  │ ?    │ GL2-W3.5      │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ *vgr │ indigo  │ 1.03    │ 1.34   │ *      │ 2.01   │ ?    │               │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ Rays │ vista   │ 116.4   │ 86.0   │ *      │ 102.3  │ ?    │ SGI UNIX      │
│      │         │         │        │        │        │      │ 3.5           │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ *vgr │ vista   │ 1.04    │ 1.35   │ *      │ 2.02   │ ?    │               │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ Rays │ virus   │ 127.3   │ 69.9   │ 57.0   │ 52.4   │ ?    │ Sun UNIX      │
│      │         │         │        │        │        │      │ 3.2           │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ *vgr │ virus   │ 1.14    │ 1.10   │ 1.04   │ 1.03   │ ?    │               │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ Rays │ vortac  │ 148.2   │ 81.0   │ 65.8   │ 60.7   │ ?    │ Sun UNIX      │
│      │         │         │        │        │        │      │ 3.2           │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ *vgr │ vortac  │ 1.32    │ 1.27   │ 1.20   │ 1.20   │ ?    │               │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ Rays │ utah-gr │ 191.8   │ 105.8  │ 89.9   │ 86.5   │ ?    │ 4.3 BSD       │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ *vgr │ utah-gr │ 1.71    │ 1.66   │ 1.64   │ 1.71   │ ?    │               │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ Rays │ ardec-3 │ -       │ 68.4   │ 57.2   │ -      │ ?    │ OSx 2.5       │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ *vgr │ ardec-3 │ -       │ 1.08   │ 1.04   │ -      │ ?    │               │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ Rays │ elxsim1 │ 380.5   │ 232.2  │ 204.8  │ 189.3  │ ?    │ BSD 4.2       │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ *vgr │ elxsim1 │ 3.39    │ 3.65   │ 3.72   │ 3.73   │ ?    │               │
├──────┼─────────┼─────────┼────────┼────────┼────────┼──────┼───────────────┤
│      │         │         │        │        │        │      │               │
│ Rays │ spark   │ 413.4   │ 232.0  │ 213.8  │ 211.9  │ ?    │ UTX 2.0       │

okay, a few things I'm obsessed with here 🫶:

  • vgr running BSD 4.3 is the whole baseline, no cap — its *vgr row is literally 1.0, 1.0, 1.0, 1.0, and it's the only row in the entire table with a Mean filled in, and that mean is 1.0. main character energy 👑.
  • Every machine gets two rows, one Rays and one *vgr, and the second set of numbers is the normalised one — much smaller, like 0.72 or 3.39.
  • indigo (GL2-W3.5) and vista (SGI UNIX 3.5) both have a * in the Star column — that's the "no data, bestie" asterisk, and their World/Bldg normalised numbers (1.34, 2.01, 1.35, 2.02) are the big ones of the group. the table says what it says 💅.
  • ardec-3 on OSx 2.5 is out here with - for Moss and Bldg — a dash, not a number, in both the Rays and the *vgr row. iconic in a chaotic way 💀.
  • violet gets two Rays rows — one plain Ultrix 1.2, one Ultrix 1.2, with w/fast_sqrt noted in the S/W column. the comma is right there in the notes, fr fr.
  • virus and vortac are both Sun UNIX 3.2, same OS, different machines — and they still don't match, which is the whole point of a benchmark, bestie ✨.
  • elxsim1 (BSD 4.2) and spark (UTX 2.0) are the slowpokes of this era — 380.5→3.39 and 413.4, which is a whole lot of seconds compared to tek4132's 80.5 🌸.
  • and yes, there's a lone .TH sitting in the A/* column of the first row, with every other cell blank. no notes. it's just there. we respect it 🙈.
  • Mean is ? for every single machine except vgr. the question marks are carrying this table ngl 😭.

the table keeps going past spark — next part picks it right back up, don't you worry 🦋.

okay so the table just KEEPS GOING and honestly? we're committed now. these are more rows of the results, same deal as before: benchmark name, machine name, those four time columns, that ? we still don't have an answer for, and whatever note the machine came with. reading them in order so nothing gets lost 💅

  • *vgr · spark — 3.69 3.65 3.89 4.18 ?
  • Rays · vmb — 413.9 233.4 210.6 212.6 ? · UTX 2.0
  • *vgr · vmb — 3.69 3.67 3.83 4.19 ?
  • Rays · c1east — 454.8 252.4 205.6 192.3 ? · Convex UNIX 6.0.1.12,
  • *vgr · c1east — 4.06 3.97 3.74 3.80 ? · vanilla cc
  • Rays · seer — 460.7 263.2 246.8 241.0 ? · UTX 2.0
  • *vgr · seer — 4.11 4.14 4.49 4.75 ?
  • Rays · venom — 492.9 228.0 180.0 157.1 ? · Concentrix 2.0
  • *vgr · venom — 4.40 3.58 3.27 3.10 ?
  • Rays · elxgnu — 520.6 315.4 264.4 242.0 ? · BSD 4.2
  • *vgr · elxgnu — 4.64 4.96 4.81 4.77 ?
  • Rays · utah-cs — 521.1 292.1 237.4 216.2 ? · BSD 4.3
  • *vgr · utah-cs — 4.64 4.59 4.32 4.27 ?
  • Rays · c1east — 521.6 285.3 230.5 215.1 ? · Convex UNIX 6.0.1.12,
  • *vgr · c1east — 4.66 4.49 4.20 4.24 ? · vc -O1 (scalar)
  • Rays · c1east — 527.7 287.4 228.7 210.0 ? · Convex UNIX 6.0.1.12,
  • *vgr · c1east — 4.71 4.52 4.16 4.14 ? · vc -O2 (vector)
  • Rays · elxgnuy — 644.7 349.3 264.6 242.1 ? · System Vr2
  • *vgr · elxgnuy — 5.75 5.49 4.81 4.77 ?
  • Rays · mflow — 845.2 439.5 313.1 338.9 ? · (preliminary)
  • *vgr · mflow — 7.54 6.91 5.69 6.68 ?
  • Rays · venom — 904.9 424.6 349.3 312.6 ? · Concentrix 2.0,
  • *vgr · venom — 8.07 6.68 6.35 6.17 ? · 2 CEs, no vectors
  • Rays · venom — 1375.5 650.0 523.4 459.9 ? · Concentrix 2.0,

and yeah, venom out here showing up three separate times because apparently one flex wasn't enough 😭 the last one's *vgr row is right on the other side of the break, so hold that thought bestie 🫶

okay so we're still in the scoreboard, bestie, and it just keeps going 💅 — so here's the rest of this stretch of rows, in exactly the order they show up. Same deal as before: type, machine, four numbers, then the ? in the fifth slot, then the little note on the end. And no cap, where you see a - it's literally a dash in the original — not a zero, not a blank space, a -. Keeping it one hundred percent faithful because that's the whole point 🫶

  • *vgr · venom — 12.27 · 10.22 · 9.52 · 9.07 · ? — 3 CEs, no vectors
  • Rays · venom — 1813.5 · 845.5 · 686.0 · 600.7 · ? — Concentrix 2.0,
  • *vgr · venom — 16.17 · 13.29 · 12.47 · 11.85 · ? — 4 CEs, no vectors
  • Rays · venom — 2364.1 · 1104.4 · 870.2 · 775.9 · ? — Concentrix 2.0,
  • *vgr · venom — 21.08 · 17.36 · 15.82 · 15.3 · ? — 5 CEs, no vectors
  • Rays · snm2 — 2492.9 · - · - · - · ? — Unicos 2.0,
  • *vgr · snm2 — 22.24 · - · - · - · ? — no vectors, no optim
  • Rays · hep — 2502.0 · - · - · - · ? — 1 PEM, npsw=10
  • *vgr · hep — 22.32 · - · - · - · ? — (no note on this one)
  • Rays · venom — 2811.4 · 1319.0 · 1051.6 · 918.0 · ? — Concentrix 2.0,
  • *vgr · venom — 25.08 · 20.74 · 19.12 · 18.1 · ? — 6 CEs, no vectors

okay and the way venom keeps showing up?? main character energy fr fr 👑 it's giving "the machine everyone had access to." love that for us.

  • Rays · venom — 3248.8 · 1533.0 · 1232.3 · 1065.9 · ? — Concentrix 2.0,
  • *vgr · venom — 28.98 · 24.10 · 22.41 · 21.02 · ? — 7 CEs, no vectors
  • Rays · patton — 3453.3 · 1514.3 · 1271.2 · 1027.3 · ? — COS V115BF2
  • *vgr · patton — 30.81 · 23.81 · 23.11 · 20.26 · ? — 1 CPU, no vectors
  • Rays · venom — 3677.3 · 1724.2 · 1375.4 · 1208.4 · ? — Concentrix 2.0,
  • *vgr · venom — 32.80 · 27.11 · 25.01 · 23.83 · ? — 8 CEs, no vectors
  • Rays · alliant — 3972.9 · 1840.0 · 1457.2 · 1266.4 · ? — Concentrix 3.0,
  • *vgr · alliant — 35.44 · 28.93 · 26.49 · 24.98 · ? — 8 CEs, no vectors
  • Rays · hep — 4055.9 · - · - · - · ? — 1 PEM, npsw=40
  • *vgr · hep — 36.18 · - · - · - · ? — (no note here either, she said nothing and left ✨)
  • Rays · patton — 6856.6 · - · 2522.2 · - · ? — COS V115BF2
  • *vgr · patton — 61.16 · - · 45.86 · - · ? — 2 CPUs, no vectors

quick vibe check on what you're looking at 🧚 notice how the *vgr rows and the Rays rows trade off, one pair at a time — a *vgr entry, then a Rays entry, then another *vgr, and so on, all the way down. And the Rays column is climbing basically the whole way down this stretch, from 1813.5 up to 6856.6 🔥 while the *vgr side is doing the same little climb in its own units, 12.27 all the way to 61.16. not me getting invested in a number going up 😭

bestie reminder: don't go "fixing" these numbers or reordering them, okay? they're printed in the order the original prints them, and a stray digit change would be a whole thing 💀 copy them exactly or don't copy them at all.

and yeah — the notes are doing a lot of the storytelling here tbh. 3 CEs, no vectors, 4 CEs, no vectors, 5 CEs, 6 CEs, 7 CEs, 8 CEs on the venom *vgr rows, 8 CEs, no vectors on alliant too, no vectors, no optim on snm2, 1 CPU, no vectors on the first patton, then 2 CPUs, no vectors on the last one. the hep rows are off doing their own thing with 1 PEM, npsw=10 and 1 PEM, npsw=40 🌸 and the Rays rows are mostly hanging out with Concentrix 2.0, / Concentrix 3.0, or COS V115BF2 — and yes the Concentrix ones really do keep that trailing comma in the original, so we keep it too 🎀

lowkey obsessed with how a table from the nineties has this much drama in it. stay tuned, there is more where this came from 💖

Okay so we're deep in the table now and honestly? The vibes are immaculate. Same machine, patton, but two different implementations doing their thing, and the numbers are not close. Let's look 👀

  • Rays on patton → 10205.1, then -, then 3749.9, then -, then ?. Machine notes: COS V115BF2 ✨
  • *vgr on patton → 91.04, then -, then 68.18, then -, then ?. Machine notes: 3 CPUs, no vectors 💀

Like… bestie. 10205.1 versus 91.04. That's not a race, that's a situation. And notice the little asterisk on *vgr? That's the mark of a certified "we tried something different here" moment, and the note literally tells you why: 3 CPUs, no vectors. So it's not even a fair fight, it's just what it is 🫶

Then we scale up:

  • Rays on patton again → 13320.2, then -, then 4955.6, then -, then ?. Machine notes: COS V115BF2 🔥
  • *vgr on patton again → 118.82, then -, then 90.10, then -, then ?. Machine notes: 4 CPUs, no vectors 👑

So more CPUs happened (3 CPUs → 4 CPUs), the *vgr numbers nudged up (91.04 → 118.82, 68.18 → 90.10), and the Rays numbers went up too (10205.1 → 13320.2, 3749.9 → 4955.6). Everything's moving in the right direction, which is the one piece of good news in this entire table ngl 😭

And those dashes? They're just dashes, okay. - means -. Some columns are filled, some aren't, and the ? at the end is doing a whole "we don't know, don't ask" main character thing 🙈

Heads up bestie: those blank-looking cells in the original are literally blank — not zero, not "failed", just nothing written there. Don't be out here reading meaning into emptiness, that's how you get got 💅

Also the COS V115BF2 note keeps showing up on every Rays row like a little signature — same machine config, both runs. Consistent queen 🌸 Note it appears only on the Rays entries in this stretch, while the *vgr entries get the CPU-count-and-no-vectors notes instead. So like… the two implementations are annotated differently, and that's just how the table is, no cap 🎀

Anyway — the takeaway from these four rows is: same box, wildly different results depending on which implementation you're looking at, and tiny CPU-count bumps do not close a gap that big. Fr fr, that's the whole story here, and it continues in the next chunk of the table 💖

Okaaay so this next block is the release 2.3 stats, dated 2-NOV-1987, and bestie it is a lot of numbers. Same column layout as before — A/*, H/W, Moss, World, Star, Bldg, Mean, S/W Notes — and the very first thing in the table is literally just the row with .TH in the A/* slot and nothing else filled in. That's the header row doing its header-row thing 💅 no numbers, no notes, just vibes.

The pattern down the whole thing: a Rays row, then a starred *vgr row right under it. Every single time. And the starred ones are the little tiny numbers (like 1.0, 3.16, 8.07) while Rays is out here in the hundreds and thousands. Not me noticing 👀 anyway, here's every row, nothing skipped, fr fr:

  • Rays · vgr — Moss 118.8, World 64.9, Star 36.5, Bldg 34.2, Mean ?. Notes: BSD 4.3
  • *vgr · vgr — Moss 1.0, World 1.0, Star 1.0, Bldg 1.0, Mean 1.0. Notes: (blank)
  • Rays · vmb — Moss 375.4, World 218.6, Star 186.4, Bldg 196.2, Mean ?. Notes: UTX 2.0
  • *vgr · vmb — Moss 3.16, World 3.37, Star 5.11, Bldg 5.74, Mean ?. Notes: (blank)
  • Rays · venom — Moss 468.2, World 221.7, Star 166.1, Bldg 144.9, Mean ?. Notes: Concentrix 3.0,
  • *vgr · venom — Moss 3.94, World 3.42, Star 4.55, Bldg 4.24, Mean ?. Notes: 1 CE, no vectors
  • Rays · sws2 — Moss 471.9, World 258.4, Star 202.5, Bldg 186.4, Mean ?. Notes: Convex Unix 6.2.32.2
  • *vgr · sws2 — Moss 3.97, World 3.98, Star 5.55, Bldg 5.45, Mean ?. Notes: vanilla cc -O
  • Rays · vector — Moss 656.6, World 301.6, Star 221.1, Bldg 195.0, Mean ?. Notes: Concentrix 3.0,
  • *vgr · vector — Moss 5.53, World 4.65, Star 6.06, Bldg 5.70, Mean ?. Notes: 1 ACE, no vectors
  • Rays · venom — Moss 827.3, World 424.0, Star 327.1, Bldg 287.9, Mean ?. Notes: Concentrix 3.0,
  • *vgr · venom — Moss 6.96, World 6.53, Star 8.96, Bldg 8.42, Mean ?. Notes: 2 CEs, no vectors
  • Rays · vhs — Moss 958.3, World 545.6, Star 447.0, Bldg 414.1, Mean ?. Notes: -O2 optimization
  • *vgr · vhs — Moss 8.07, World 9.41, Star 12.25, Bldg 12.11, Mean ?. Notes: (blank)
  • Rays · mseries — Moss 988.3, World 603.5, Star 515.6, Bldg 491.6, Mean ?. Notes: (blank)
  • *vgr · mseries — Moss 8.32, World 9.29, Star 14.13, Bldg 14.37, Mean ?. Notes: (blank)
  • Rays · vector — Moss 1280.9, World 596.8, Star 436.7, Bldg 389.2, Mean ?. Notes: Concentrix 3.0,
  • *vgr · vector — Moss 10.78, World 9.19, Star 11.96, Bldg 11.38, Mean ?. Notes: 2 ACEs, no vectors
  • Rays · venom — Moss 1288.7, World 630.6, Star 486.4, Bldg 428.0, Mean ?. Notes: Concentrix 3.0,
  • *vgr · venom — Moss 10.85, World 9.72, Star 13.33, Bldg 12.51, Mean ?. Notes: 3 CEs, no vectors
  • Rays · venom — Moss 1732.9, World 853.2, Star 645.4, Bldg 560.1, Mean ?. Notes: Concentrix 3.0,

a couple things my brain refuses to shut up about 🧚:

  • That one *vgr / vgr row is the only line in the whole block where Mean isn't a ? — it's 1.0, and every other column on that row is 1.0 too. Iconic. Slay. Baseline behavior.
  • The S/W Notes column is out here doing the most: BSD 4.3, UTX 2.0, Convex Unix 6.2.32.2, a whole bunch of Concentrix 3.0, (with the trailing comma just… sitting there, love that for it), vanilla cc -O, -O2 optimization, and the little vector diary entries: 1 CE, no vectors, 1 ACE, no vectors, 2 CEs, no vectors, 2 ACEs, no vectors, 3 CEs, no vectors. The escalation is giving character development 💀
  • Some rows genuinely have nothing in the notes column, and that's fine, we don't judge.
Heads up bestie: the Mean column is ? on basically every row except that 1.0 one — that's straight from the page, not me being mysterious 🫶

and we're still mid-table, so stay tuned 🦋

okay bestie we're still deep in the numbers trenches 😭 but like — this is the good part, this is where the Rays numbers and the *vgr numbers are just trading blows like it's a tennis match, and honestly? 🎾

same deal as the last batch: the four number columns, then the ? column, then the notes. zero edits, zero rounding, I'm not touching a single digit fr fr 💅

  • *vgr · venom — 14.59, 13.15, 17.68, 16.38 · ? · 4 CEs, no vectors
  • Rays · bob — 1856.7, 801.9, 574.1, 500.3 · ? · 1 CPU, UNICOS 3.0,
  • *vgr · bob — 15.63, 12.36, 15.73, 15.62 · ? · CC -O

look at bob showing up with a whole real-machine score AND then a virtual one right after. main character behavior, we love to see it 👑

  • Rays · vector — 1952.1, 881.7, 647.6, 578.8 · ? · Concentrix 3.0,
  • *vgr · vector — 16.43, 13.58, 17.74, 16.92 · ? · 3 ACEs, no vectors
  • Rays · venom — 2158.4, 1031.3, 797.3, 703.1 · ? · Concentrix 3.0,
  • *vgr · venom — 18.17, 15.89, 21.84, 20.56 · ? · 5 CEs, no vectors
  • Rays · bob — 2346.1, 1127.4, 751.0, 677.2 · ? · 1 CPU UNICOS 4.0
  • *vgr · bob — 19.75, 17.37, 20.57, 19.80 · ? · CC -O
  • Rays · bob — 2423.5, 1045.7, 719.3, 609.4 · ? · 1 CPU, UNICOS 3.0,
  • *vgr · bob — 20.40, 16.11, 19.71, 17.82 · ? · VCC -O

okay the bob entries are stacking up like a whole little subplot 📚 and notice how we're hopping between CC -O and VCC -O — the compiler pipeline is doing a whole thing and I am watching.

  • Rays · vector — 2546.2, 1170.0, 867.2, 758.7 · ? · Concentrix 3.0,
  • *vgr · vector — 21.43, 18.03, 23.76, 22.18 · ? · 4 ACEs, no vectors
  • Rays · venom — 2633.7, 1232.7, 943.4, 839.5 · ? · Concentrix 3.0,
  • *vgr · venom — 22.17, 18.99, 25.85, 24.55 · ? · 6 CEs, no vectors
  • Rays · venom — 2769.4, 1393.3, 1098.1, 971.6 · ? · Concentrix 3.0,
  • *vgr · venom — 23.31, 21.47, 30.08, 28.41 · ? · 7 CEs, no vectors
  • Rays · vector — 3131.2, 1453.8, 1065.8, 945.4 · ? · Concentrix 3.0,
  • *vgr · vector — 26.36, 22.40, 29.20, 27.64 · ? · 5 ACEs, no vectors
  • Rays · venom — 3368.7, 1605.3, 1231.3, 1095.11 · ? · Concentrix 3.0,
  • *vgr · venom — 28.36, 24.73, 33.73, 32.02 · ? · 8 CEs, no vectors
yes 1095.11 is really in there with the two decimal places, and no I'm not "fixing" it 💀 what's in the table stays in the table, that's the whole entire rule.

and then patton rolls in at the end like it heard there was a party 🎉

  • Rays · patton — 3456.5, 1691.8, 1169.0, 1043.6 · ? · 1 CPU, UNICOS 2.1

that's the batch, and there's more where that came from — do not close the tab bestie, we're literally mid-table 🌸

okay bestie, we're mid-table and the rows just keep coming, so let's keep scrolling together 💅 every row below is exactly as it appears — same order, same numbers, nothing skipped, nothing rounded. The four numeric columns come in the same order every time, then there's that ? column that is literally just a question mark in every single row (yes really, that's the value, don't come for me 😭), and then the trailing note column. Buckle up 🎀

  • *vgr on patton — 29.10 · 26.07 · 32.03 · 30.51 — ? — note: no vectors
  • Rays on vector — 3686.4 · 1688.6 · 1259.1 · 1129.0 — ? — note: Concentrix 3.0,
  • *vgr on vector — 31.03 · 26.02 · 34.49 · 33.01 — ? — note: 6 ACEs, no vectors
  • Rays on vector — 4140.3 · 1953.0 · 1463.3 · 1299.7 — ? — note: Concentrix 3.0,
  • *vgr on vector — 34.85 · 30.09 · 40.09 · 38.00 — ? — note: 7 ACEs, no vectors
  • Rays on bob — 4641.7 · 2243.4 · 1498.2 · 1358.1 — ? — note: 2 CPUs, UNICOS 4.0
  • *vgr on bob — 39.07 · 34.57 · 41.05 · 39.71 — ? — note: CC -O
  • Rays on vector — 4712.2 · 2226.3 · 1649.0 · 1469.7 — ? — note: Concentrix 3.0,
  • *vgr on vector — 39.66 · 34.30 · 45.18 · 42.97 — ? — note: 8 ACEs, no vectors
  • Rays on bob — 6884.9 · 3422.4 · 2267.7 · 2034.3 — ? — note: 3 CPUs, UNICOS 4.0
  • *vgr on bob — 57.95 · 52.73 · 62.13 · 59.48 — ? — note: CC -O
  • Rays on bob — 9477.1 · 4484.7 · 2947.0 · 2832.9 — ? — note: 4 CPUs, UNICOS 4.0
  • *vgr on bob — 79.77 · 69.10 · 80.76 · 82.83 — ? — note: CC -O

a couple of things I need you to notice, because they're iconic ✨

first: the Concentrix 3.0, notes end in a comma, fr fr. That's not me being messy — the note is literally written that way in the page, trailing comma and all, so I'm keeping it exactly as-is 🙈

second: the vibe check on this chunk of rows — the Rays entries are all huge four-figure-ish numbers like 3686.4 and 9477.1, while every *vgr entry is a small number in the tens like 29.10 and 79.77. Same table, wildly different magnitude, and the note column is always there telling you what box it ran on: no vectors, 6 ACEs, no vectors, 7 ACEs, no vectors, 8 ACEs, no vectors, 2 CPUs, UNICOS 4.0 all the way up to 4 CPUs, UNICOS 4.0, and plain old CC -O for the *vgr side 💀

and the number of ACEs just keeps climbing as the rows go down: none → 6 → 7 → 8. little glow-up arc, honestly 👑

more rows incoming, don't go anywhere 🦋

🎀 RELEASE 3.0 STATISTICS (10-OCT-1988) 🎀

Okay bestie, this is the 3.0 era and the table is giving full spreadsheet realness 😭 — eight columns, A/*, H/W, Moss, World, Star, Bldg, Mean, and S/W Notes. Every machine gets two rows: an Abs row with the absolute numbers, and a *vgr row right under it. And the very first row is just .TH sitting in the A/* column with everything else blank — iconic header behavior, no notes 💅.

Also: vgr is the one whose *vgr row is all 1.00 across the board. Baseline energy, periodt ✨.

🖥️ the actual lineup

  • vgr — H/W vgr
    • Abs → Moss 138.85 · World 67.15 · Star 54.48 · Bldg 49.11 · Mean 77.39 · S/W: BSD 4.3
    • *vgr → Moss 1.00 · World 1.00 · Star 1.00 · Bldg 1.00 · Mean 1.00 · S/W: w/FPA
  • sdm — H/W sdm
    • Abs → Moss 163.56 · World 76.98 · Star 59.79 · Bldg 52.51 · Mean 88.21 · S/W: SunOS 3.4
    • *vgr → Moss 1.17 · World 1.14 · Star 1.09 · Bldg 1.06 · Mean 1.11 · S/W: w/68881
  • ardec-3 — H/W ardec-3
    • Abs → Moss 417.95 · World 212.43 · Star 177.01 · Bldg 160.73 · Mean 242.03 · S/W: 9820, OSx 4.1
    • *vgr → Moss 3.01 · World 3.16 · Star 3.24 · Bldg 3.27 · Mean 3.17 · S/W: -P 1
  • spark — H/W spark
    • Abs → Moss 443.84 · World 216.92 · Star 190.42 · Bldg 192.82 · Mean 261.00 · S/W: UTX 2.0
    • *vgr → Moss 3.19 · World 3.23 · Star 3.49 · Bldg 3.92 · Mean 3.37 · S/W: no MACC
  • vmb — H/W vmb
    • Abs → Moss 482.28 · World 226.33 · Star 191.52 · Bldg 193.04 · Mean 273.29 · S/W: UTX 2.0
    • *vgr → Moss 3.47 · World 3.37 · Star 3.51 · Bldg 3.93 · Mean 3.57 · S/W: no MACC
  • sws2 — H/W sws2
    • Abs → Moss 546.23 · World 255.45 · Star 183.10 · Bldg 163.85 · Mean 287.15 · S/W: Convex 6.2.32.2
    • *vgr → Moss 3.93 · World 3.80 · Star 3.36 · Bldg 3.33 · Mean 3.60 · S/W: /bin/cc
  • sun4 — H/W sun4
    • Abs → Moss 767.7 · World 373.7 · Star 315.5 · Bldg 284.5 · Mean 435.6 · S/W: Unix 3.5
    • *vgr → Moss 5.52 · World 5.56 · Star 5.79 · Bldg 5.79 · Mean 5.62 · S/W: -O2 optim
    • (yes the Abs numbers here have one decimal place and not two — that's exactly how they came, don't @ me 🙈)
  • video — H/W video
    • Abs → Moss 940.33 · World 469.30 · Star 399.21 · Bldg 367.64 · Mean 544.12 · S/W: UNIX 2.0
    • *vgr → Moss 6.77 · World 6.98 · Star 7.32 · Bldg 7.48 · Mean 7.13 · S/W: -O2
  • vhs — H/W vhs
    • Abs → Moss 954.81 · World 471.90 · Star 409.78 · Bldg 392.55 · Mean 557.26 · S/W: IRix 3.1
    • *vgr → Moss 6.87 · World 7.02 · Star 7.52 · Bldg 7.99 · Mean 7.45 · S/W: Parallel, -P1
  • bob — H/W bob 💀 (yes, bob)
    • Abs → Moss 2602.68 · World 1078.25 · Star 712.39 · Bldg 642.42 · Mean 1258.93 · S/W: UNICOS 4.0, /bin/cc
    • *vgr → Moss 21.22 · World 18.17 · Star 15.39 · Bldg 13.98 · Mean 17.19 · S/W: -P 1
  • vector — H/W vector 🔥
    • Abs → Moss 3811.68 · World 1469.80 · Star 1060.50 · Bldg 921.21 · Mean 1815.79 · S/W: Alliant 3.0
    • *vgr → Moss 27.45 · World 21.88 · Star 19.46 · Bldg 18.75 · Mean 21.88 · S/W: -P 5
    • ngl the World and Mean both landing on 21.88 is a whole mood 😌

Big numbers, tiny numbers, 1.00s everywhere for vgr — 3.0 said what it said. 💖

There's so much more where this came from, keep scrolling bestie 🫶

Omg okay, the table continues — and bestie, these rows are sorted by that first timing column going up and up and up 📈. Every row below is one entry, Abs or *vgr in the first column, the variant name next, then those five timing columns in order, then whatever system/flags were noted. Numbers stay exactly as they are, no touching 💅

  • Abs vector — 4515.09 · 1751.94 · 1266.49 · 1090.85 · 2156.09 — Alliant 3.0
  • *vgr vector — 32.51 · 26.08 · 23.24 · 22.21 · 26.01 — -P 6
  • Abs patton — 4550.67 · 1920.15 · 1298.61 · 1153.02 · 2230.61 — UNICOS 3.0, /bin/cc
  • *vgr patton — 37.10 · 32.36 · 28.05 · 25.09 · 30.65 — -P 1
  • Abs vector — 5130.74 · 2004.49 · 1438.97 · 1261.36 · 2458.89 — Alliant 3.0
  • *vgr vector — 36.95 · 29.85 · 26.41 · 25.68 · 29.72 — -P 7
  • Abs bob — 5186.24 · 2132.17 · 1409.10 · 1273.72 · 2500.30 — UNICOS 4.0, /bin/cc
  • *vgr bob — 42.28 · 35.93 · 30.44 · 27.71 · 34.09 — -P 2
  • Abs vector — 5813.76 · 2260.61 · 1638.31 · 1422.82 · 2783.87 — Alliant 3.0
  • *vgr vector — 41.87 · 33.66 · 30.07 · 28.97 · 33.64 — -P 8
  • Abs amber — 5901.29 · 2702.65 · 2096.35 · 1893.41 · 3148.42 — (the system/flag field on this one is literally blank in the original, no cap 🙈)
  • *vgr amber — 42.50 · 40.24 · 38.47 · 38.55 · 39.94 — -P 1
  • Abs bob — 7934.12 · 3392.98 · 2119.11 · 1929.74 · 3843.98 — UNICOS 4.0, /bin/cc
  • *vgr bob — 64.69 · 57.18 · 45.78 · 41.99 · 52.41 — -P 3
  • Abs patton — 8898.79 · 3782.55 · 2559.48 · 2292.13 · 4383.23 — UNICOS 3.0, /bin/cc
  • *vgr patton — 72.56 · 63.75 · 55.30 · 49.88 · 60.37 — -P 2
  • Abs bob — 10734.52 · 4436.61 · 2883.53 · 2617.47 · 5168.03 — UNICOS 4.0, /bin/cc
  • *vgr bob — 87.52 · 74.77 · 62.30 · 56.96 · 70.38 — -P 4
  • Abs patton — 13078.79 · 5631.95 · 3832.88 · 3416.90 · 6490.13 — UNICOS 3.0, /bin/cc
  • *vgr patton — 106.64 · 94.92 · 82.81 · 74.36 · 89.68 — -P 3
  • Abs patton — 17157.78 · 7437.55 · 5073.58 · 4531.25 · 8550.04 — UNICOS 3.0, /bin/cc
Notice how the patton / bob / vector names keep swapping around as the numbers climb? Yeah, they're interleaved on purpose — every entry owns its own numbers, so don't go blending them into one vibe check, bestie 🫶 And fr, don't retype these digits from memory. Copy them. Every decimal place is doing work ✨

More rows coming, don't go anywhere 💖

       │ *vgr │ patton  │ 139.90   │ 125.35  │ 109.62  │ 98.61   │ 118.37  │ -P 4      │
       └──────┴─────────┴──────────┴─────────┴─────────┴─────────┴─────────┴───────────┘

And that's the table, bestie — box closed, borders drawn, we are done 💅

Last row goes to *vgr running on patton, and yes that little asterisk is literally in the original, I'm not out here inventing punctuation, that's not the vibe 🙈 The numbers land in the exact same column order as every row above it: 139.90, 125.35, 109.62, 98.61, 118.37 — and the whole run went down with -P 4 on the end.

Not me getting emotionally attached to a row with an asterisk in it 😭 but honestly? iconic. She closed out the table, she said what she said, -P 4 and all. Periodt. 🎀

Real talk though: I'm not gonna sit here and guess what each of those five columns measures — the headings are up top in the part you already read, and I refuse to make stuff up about benchmark numbers, that's how you get clowned in the group chat 💀 Go check the header row, iykyk.

🌸 RELEASE 3.7 STATISTICS (19-JUNE-1989) 🌸

okay so this is the benchmark scrapbook page from 19-JUNE-1989, and it is literally giving vintage spreadsheet realness ✨ The column lineup is: A/*, H/W, Moss, World, Star, Bldg, M35, Mean, CPU.

There’s a .TH entry chillin’ all alone in the A/* column, with every other column blank. iconic mysterious behavior, no cap 💅

For each benchmark below, both the Abs row and the *vgr row are present. The H/W value is the benchmark name. The *vgr rows have a blank CPU column — that’s just how the table is set up, bestie 🫶

  • vgr — H/W: vgr; CPU: 1
    • Abs — Moss: 128.64, World: 54.47, Star: 45.17, Bldg: 40.35, M35: 49.02, Mean: 63.53
    • *vgr — Moss: 1.00, World: 1.00, Star: 1.00, Bldg: 1.00, M35: 1.00, Mean: 1.00; CPU: (blank)
  • rh2 — H/W: rh2; CPU: 1
    • Abs — Moss: 1704.56, World: 791.31, Star: 599.81, Bldg: 549.74, M35: 589.01, Mean: 846.88
    • *vgr — Moss: 13.25, World: 14.52, Star: 13.27, Bldg: 13.62, M35: 12.01, Mean: 13.33; CPU: (blank)
  • spark — H/W: spark; CPU: 1
    • Abs — Moss: 435.69, World: 208.59, Star: 176.76, Bldg: 177.69, M35: 174.67, Mean: 234.68
    • *vgr — Moss: 3.53, World: 3.61, Star: 4.17, Bldg: 4.29, M35: 3.40, Mean: 3.80; CPU: (blank)
  • vhs — H/W: vhs; CPU: 1
    • Abs — Moss: 973.11, World: 475.00, Star: 384.62, Bldg: 351.12, M35: 343.44, Mean: 505.45
    • *vgr — Moss: 7.56, World: 8.72, Star: 8.51, Bldg: 8.70, M35: 7.00, Mean: 8.09; CPU: (blank)
  • whisper — H/W: whisper; CPU: 1
    • Abs — Moss: 1118.35, World: 561.81, Star: 473.52, Bldg: 439.76, M35: 509.26, Mean: 620.54
    • *vgr — Moss: 8.69, World: 10.31, Star: 10.48, Bldg: 10.89, M35: 10.38, Mean: 10.15; CPU: (blank)
  • nova — H/W: nova; CPU: 1
    • Abs — Moss: 2024.80, World: 976.01, Star: 790.17, Bldg: 732.35, M35: ?, Mean: 1130.83
    • *vgr — Moss: 15.74, World: 17.92, Star: 17.49, Bldg: 18.15, M35: ?, Mean: 17.32; CPU: (blank)
not the M35 cell for nova being a whole literal ? 😭 the table said what it said, bestie
  • ovoyage — H/W: ovoyage; CPU: 2
    • Abs — Moss: 1969.57, World: 1001.21, Star: 862.77, Bldg: 809.61, M35: 786.11, Mean: 1085.85
    • *vgr — Moss: 15.31, World: 18.38, Star: 19.10, Bldg: 20.06, M35: 16.03, Mean: 17.77; CPU: (blank)
  • taylor — H/W: taylor; CPU: 1
    • Abs — Moss: 2446.98, World: 1076.24, Star: 912.03, Bldg: 837.56, M35: 799.58, Mean: 1214.47
    • *vgr — Moss: 19.02, World: 19.75, Star: 20.19, Bldg: 20.75, M35: 16.31, Mean: 19.20; CPU: (blank)
  • worm — H/W: worm; CPU: 1
    • Abs — Moss: 2671.46, World: 1218.49, Star: 966.76, Bldg: 859.61, M35: 932.18, Mean: 1329.70
    • *vgr — Moss: 20.76, World: 22.36, Star: 21.40, Bldg: 21.30, M35: 19.01, Mean: 20.96; CPU: (blank)
  • viper — H/W: viper; CPU: 8
    • Abs — Moss: 3255.73, World: 1368.96, Star: 1102.71, Bldg: 1039.37, M35: 1008.15, Mean: 1554.98
    • *vgr — Moss: 25.30, World: 25.13, Star: 24.41, Bldg: 25.75, M35: 20.56, Mean: 24.23; CPU: (blank)
  • vector — H/W: vector; CPU: 8
    • Abs — Moss: 3956.86, World: 1625.23, Star: 1234.03, Bldg: 1137.36, M35: 1121.41, Mean: 1814.97
    • *vgr — Moss: 30.75, World: 29.83, Star: 27.31, Bldg: 28.18, M35: 22.87, Mean: 27.78; CPU: (blank)

okay bestie, this next chunk is pure numbers and she is dense — so I'm dropping the whole block verbatim first (nothing lost, nothing rounded, every digit exactly where she belongs) and then we'll yap about it 🫶

       │      │         │         │         │         │         │         │         │     │
       │ Abs  │ taylor  │ 4730.05 │ 2225.56 │ 1742.09 │ 1590.57 │ 1551.78 │ 2368.01 │ 2   │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ *vgr │ taylor  │ 36.76   │ 40.85   │ 38.56   │ 39.41   │ 31.65   │ 37.44   │     │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ Abs  │ worm    │ 5017.85 │ 2309.86 │ 1894.92 │ 1708.27 │ 1791.11 │ 2544.40 │ 2   │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ *vgr │ worm    │ 39.00   │ 42.40   │ 41.95   │ 42.33   │ 36.53   │ 40.44   │     │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ Abs  │ worm    │ 7082.79 │ 3359.99 │ 2740.50 │ 2502.03 │ 2594.37 │ 3655.93 │ 3   │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ *vgr │ worm    │ 55.05   │ 61.68   │ 60.67   │ 62.00   │ 52.92   │ 58.46   │     │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ Abs  │ worm    │ 8664.78 │ 4162.53 │ 3323.46 │ 3115.36 │ 3378.75 │ 4528.97 │ 4   │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ *vgr │ worm    │ 67.35   │ 76.41   │ 73.57   │ 77.20   │ 68.92   │ 72.69   │     │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ Abs  │ worm    │ 10072.1 │ 5016.57 │ 4032.60 │ 3855.90 │ 4043.11 │ 5404.06 │ 5   │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ *vgr │ worm    │ 78.29   │ 92.09   │ 89.27   │ 95.56   │ 82.47   │ 87.53   │     │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ Abs  │ worm    │ 11678.4 │ 5752.20 │ 4758.80 │ 4493.16 │ 4700.52 │ 6276.60 │ 6   │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ *vgr │ worm    │ 90.78   │ 105.60  │ 105.35  │ 111.35  │ 95.88   │ 101.79  │     │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ Abs  │ worm    │ 12282.3 │ 6021.31 │ 5394.36 │ 4908.98 │ 5141.71 │ 6749.72 │ 7   │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ *vgr │ worm    │ 95.47   │ 110.54  │ 119.42  │ 121.65  │ 104.89  │ 110.39  │     │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ Abs  │ worm    │ 12158.3 │ 5913.50 │ 5784.67 │ 5327.66 │ 5505.71 │ 6937.96 │ 8   │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ *vgr │ worm    │ 94.51   │ 108.56  │ 128.06  │ 132.03  │ 112.31  │ 115.09  │     │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ Abs  │ patton  │ 16028.8 │ 7289.7  │ 4827.9  │ 4392.5  │ 20829.2 │ 53368.1 │ 4   │
       ├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────┤
       │      │         │         │         │         │         │         │         │     │
       │ *vgr │ patton  │ 124.6   │ 133.83  │ 106.88  │ 108.85  │ 106.23  │ 116.07  │     │
       └──────┴─────────┴─────────┴─────────┴─────────┴─────────┴─────────┴─────────┴─────┘

so the way to read her is: first column is the row type — Abs for the absolute number, *vgr for the ratio one. second column is which program it was: taylor, worm, or patton. then the six machine columns (left to right, in the order the header up above laid out), and that lil trailing digit column doing her own thing 💅

🌸 the receipts, program by program 🌸

  • taylor — one Abs row, one *vgr row. The Abs line hits 4730.05, 2225.56, 1742.09, 1590.57, 1551.78, 2368.01 across the six machines and carries a trailing 2. The *vgr line comes in way lower: 36.76, 40.85, 38.56, 39.41, 31.65, 37.44. She's the calm one in this whole table, ngl 😌
  • worm — okay this is the main character of the block, seven Abs/*vgr pairs and she climbs. The Abs values go 5017.85, 2309.86, 1894.92, 1708.27, 1791.11, 2544.40 (trailing 2), then 7082.79, 3359.99, 2740.50, 2502.03, 2594.37, 3655.93 (trailing 3), then 8664.78, 4162.53, 3323.46, 3115.36, 3378.75, 4528.97 (trailing 4), then 10072.1, 5016.57, 4032.60, 3855.90, 4043.11, 5404.06 (trailing 5), then 11678.4, 5752.20, 4758.80, 4493.16, 4700.52, 6276.60 (trailing 6), then 12282.3, 6021.31, 5394.36, 4908.98, 5141.71, 6749.72 (trailing 7), and finally 12158.3, 5913.50, 5784.67, 5327.66, 5505.71, 6937.96 (trailing 8) — yes bestie, the last one dips slightly instead of rising, the way that not every curve is a perfect line, we love an honest data point 🙈
  • worm *vgr rows, in the same order — 39.00, 42.40, 41.95, 42.33, 36.53, 40.44 → 55.05, 61.68, 60.67, 62.00, 52.92, 58.46 → 67.35, 76.41, 73.57, 77.20, 68.92, 72.69 → 78.29, 92.09, 89.27, 95.56, 82.47, 87.53 → 90.78, 105.60, 105.35, 111.35, 95.88, 101.79 → 95.47, 110.54, 119.42, 121.65, 104.89, 110.39 → 94.51, 108.56, 128.06, 132.03, 112.31, 115.09. same story, different outfit: up and up, then a tiny exhale at the end 💜
  • patton — one Abs row and one *vgr row, and the Abs line is unhinged in the best way: 16028.8, 7289.7, 4827.9, 4392.5, 20829.2, 53368.1 with a trailing 4. look at that last column. look at her. she said "no thoughts, just a number so big it needs its own zip code" 🔥
  • patton *vgr — 124.6, 133.83, 106.88, 108.85, 106.23, 116.07. comparatively chill, which is so funny given what the Abs row just did 💀

a couple things I want you to hold in your brain while you scroll, bestie:

the Abs rows and the *vgr rows are two different beasts and you cannot compare across them — one is absolute, one is a ratio, and mixing them is how you end up looking silly in the group chat. read each row in her own lane 🫶
the trailing digit column only shows up on the Abs rows here, and she walks 2, 2, 3, 4, 5, 6, 7, 8, then 4. I am not inventing meaning for her, I'm just reporting what she said ✨

and yeah — worm showing up seven separate times with a rising number is lowkey the whole vibe of this block. it's giving "let me cook" 👑

more coming, don't touch that scroll wheel 🎀

🌸 RELEASE 4.0 STATISTICS 🌸

okay bestie, this one's a chunky one — it's the 4.0 numbers and they're giving spreadsheet era 😭 the original prints this as one giant box-drawing table, and we're gonna walk through it row by row so nothing gets lost, no cap.

first, the vibe check on the columns, because the header row is doing the most:

  • A/* — tells you whether the row is absolute (Abs) or the ratio against vgr (*vgr)
  • H/W — the test name (i.e. which hardware test we're looking at)
  • the machine columns, in order: Moss, World, Star, Bldg, M35, then the Mean, and finally CPU

and yeah, the literal header row in the original has a .TH sitting in the H/W column — that's just the manpage macro peeking through, she's shy 🙈 ignore her.

one more thing before we go: slc1 genuinely has a ? in the M35 column in the original. Not me losing your data — that's literally what's printed 💀

the tests, one by one ✨

  • vgr
    • Abs — Moss 137.82 · World 67.23 · Star 56.39 · Bldg 53.91 · M35 69.25 · Mean 76.92 · CPU 1
    • *vgr — Moss 1.00 · World 1.00 · Star 1.00 · Bldg 1.00 · M35 1.00 · Mean 1.00
  • slc1
    • Abs — Moss 153.63 · World 80.95 · Star 62.94 · Bldg 60.48 · M35 ? · Mean 71.60 · CPU 1
    • *vgr — Moss 1.12 · World 1.20 · Star 1.12 · Bldg 1.13 · M35 ? · Mean .91
  • vision
    • Abs — Moss 165.85 · World 89.78 · Star 76.44 · Bldg 75.78 · M35 96.18 · Mean 100.80 · CPU 1
    • *vgr — Moss 1.21 · World 1.33 · Star 1.36 · Bldg 1.42 · M35 1.36 · Mean 1.33
  • vmb
    • Abs — Moss 381.92 · World 195.11 · Star 178.78 · Bldg 186.62 · M35 214.31 · Mean 231.34 · CPU 1
    • *vgr — Moss 2.78 · World 2.90 · Star 3.18 · Bldg 3.49 · M35 3.03 · Mean 3.07
  • sws1
    • Abs — Moss 636.26 · World 288.76 · Star 212.99 · Bldg 210.18 · M35 271.02 · Mean 323.84 · CPU 1
    • *vgr — Moss 4.64 · World 4.30 · Star 3.79 · Bldg 3.93 · M35 3.83 · Mean 4.09
  • veto
    • Abs — Moss 827.25 · World 436.38 · Star 361.89 · Bldg 348.99 · M35 422.19 · Mean 479.34 · CPU 1
    • *vgr — Moss 6.03 · World 6.50 · Star 6.45 · Bldg 6.54 · M35 5.97 · Mean 6.29
  • vhs
    • Abs — Moss 1040.44 · World 500.52 · Star 436.90 · Bldg 415.15 · M35 496.30 · Mean 577.86 · CPU 1
    • *vgr — Moss 7.59 · World 7.46 · Star 7.79 · Bldg 7.78 · M35 7.02 · Mean 7.52
  • cor3
    • Abs — Moss 1203.14 · World 553.22 · Star 445.45 · Bldg 398.79 · M35 534.18 · Mean 626.95 · CPU 1
    • *vgr — Moss 8.78 · World 8.24 · Star 7.94 · Bldg 7.47 · M35 7.55 · Mean 7.99
  • wilted
    • Abs — Moss 1556.66 · World 800.18 · Star 722.95 · Bldg 708.21 · M35 858.90 · Mean 929.38 · CPU 1
    • *vgr — Moss 11.36 · World 11.93 · Star 12.89 · Bldg 13.27 · M35 12.15 · Mean 12.32
  • shpcrc2
    • Abs — Moss 2274.36 · World 1041.79 · Star 915.99 · Bldg 854.30 · M35 1110.23 · Mean 1239.33 · CPU 1
    • *vgr — Moss 16.59 · World 15.53 · Star 16.33 · Bldg 16.01 · M35 15.70 · Mean 16.03
  • walrus
    • Abs — Moss 2291.38 · World 1094.04 · Star 960.07 · Bldg 892.67 · M35 1066.31 · Mean 1260.89 · CPU 1
    • *vgr — Moss 16.72 · World 16.31 · Star 17.12 · Bldg 16.72 · M35 15.08 · Mean 16.39

quick note on how to read this, because the *vgr rows are the whole point 👑 — every one of them is scaled against vgr, which is why vgr's own *vgr row is 1.00 across the board. the Abs rows are the raw ones, and the CPU column is just 1 down the entire thing because this is the baseline of that column. the Mean column pulls together everything to the left of it for that row, so walrus at Abs Mean 1260.89 and *vgr Mean 16.39 is the last line of the table — that's where 4.0 signs off ✨

okay bestie the table literally will not stop and we are NOT leaving a single digit behind 💅 so here's the rest of it, in exactly the order it shows up — Abs row, then its *vgr row, rinse and repeat — and note that the last column has a number on the Abs rows and is just gonna sit there empty on the *vgr rows 🫶

  • Abs walrus — 4010.55, 2019.34, 1767.75, 1685.19, 2029.78, 2302.52 — last column: 2
  • *vgr walrus — 29.27, 30.11, 31.52, 31.58, 28.71, 30.23 — last column: empty

next up, a whole new name, okay we love a fresh entry ✨

  • Abs patton — 5235.87, 2108.02, 1509.85, 1507.81, 2132.01, 2498.71 — last column: 1
  • *vgr patton — 38.21, 31.43, 26.92, 28.25, 30.15, 30.99 — last column: empty
  • Abs crim — 4679.97, 2246.04, 2143.07, 1936.85, 2224.65, 2646.11 — last column: 1
  • *vgr crim — 34.15, 33.49, 38.22, 36.29, 31.47, 34.72 — last column: empty
  • Abs uy1 — 6220.79, 2474.97, 1742.33, 1748.73, 2464.90, 2930.34 — last column: 1
  • *vgr uy1 — 45.40, 36.90, 31.07, 32.77, 34.86, 36.20 — last column: empty

and here's where a flag finally shows up in that last column, so eyes on it bestie 👀

  • Abs wiltse — 5785.83, 2737.06, 2473.66, 2310.58, 2699.06, 3201.23 — last column: 1
  • *vgr wiltse — 42.22, 40.81, 44.11, 43.30, 38.18, 41.72 — last column: -mips1
  • Abs walrus — 5743.85, 2938.73, 2513.24, 2446.85, 2858.84, 3300.30 — last column: 3
  • *vgr walrus — 41.92, 43.82, 44.82, 45.85, 40.44, 43.37 — last column: empty
  • Abs wiltse — 7093.51, 3458.89, 2887.77, 2730.62, 3199.08, 3873.97 — last column: 1
  • *vgr wiltse — 51.77, 51.57, 51.50, 51.17, 45.25, 50.25 — last column: -mips2

okay and now the numbers start doing the absolute most, like the way these climb?? we're obsessed 🔥

  • Abs walrus — 7131.48, 3613.02, 3262.01, 3048.77, 3594.35, 4129.92 — last column: 4
  • *vgr walrus — 52.05, 53.87, 58.17, 57.13, 50.84, 54.41 — last column: empty
  • Abs whiz — 7159.01, 3655.65, 3355.85, 3132.36, 3655.02, 4191.57 — last column: 4
  • *vgr whiz — 52.25, 54.51, 59.85, 58.70, 51.70, 55.40 — last column: empty
  • Abs walrus — 7698.43, 4119.06, 3889.37, 3690.27, 4244.30, 4728.28 — last column: 5
  • *vgr walrus — 56.18, 61.42, 69.36, 69.15, 60.04, 63.23 — last column: empty
  • Abs walrus — 8718.08, 4625.56, 4469.49, 4295.53, 4765.40, 5374.81 — last column: 6
  • *vgr walrus — 63.63, 68.97, 79.71, 80.50, 67.41, 72.04 — last column: empty
  • Abs walrus — 8937.71, 4882.66, 4911.99, 4601.09, 5335.14, 5733.71 — last column: 7
  • *vgr walrus — 65.23, 72.81, 87.60, 86.22, 75.47, 77.46 — last column: empty
real talk: every single number above is copied exactly, no rounding, no "close enough", no manifesting different digits into existence 🙈 if a value looks cursed, it's cursed on purpose and it's staying that way.

and yeah — walrus is carrying this whole stretch of the table on its back, main character energy fr fr 👑

├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼────────┤
│      │         │         │         │         │         │         │         │        │
│ Abs  │ walrus  │ 9300.31 │ 5136.17 │ 5243.14 │ 5112.23 │ 5318.45 │ 6022.06 │ 8      │
├──────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼─────────┼────────┤
│      │         │         │         │         │         │         │         │        │
│ *vgr │ walrus  │ 67.88   │ 76.59   │ 93.51   │ 95.80   │ 75.23   │ 81.80   │        │
└──────┴─────────┴─────────┴─────────┴─────────┴─────────┴─────────┴─────────┴────────┘

…and that's the last of the numbers, bestie, we survived the tables 🫶 (no cap, the absolute and *vgr rows hitting differently on the same walrus box is so iconic — look at that spread, it's giving "same machine, wildly different energy" 💀).

✨ ADDITIONAL STATISTICS ✨

okay so this is the part where we ask a favour, and it's genuinely the easiest good deed in all of Unix 🌸

The BRL-CAD developers would be greatly appreciative if you'd e-mail them any statistics you gather. Like, please — they're literally collecting these things by hand, be the bestie who sends the data 💖.

When you do, here's the whole wishlist of what to include — don't skip any of it, every single line helps them make sense of your numbers:

  • the "summary" file produced by the benchmark script 📄 (yes, the whole file, that's the main character of the submission)
  • which BRL-CAD release was used
  • the manufacturer of the machine
  • the hardware model name/number
  • the software version numbers (OS and compiler)
  • the presence or absence of floating point hardware
  • processor cache sizes
  • bus and memory speeds
  • the local host name (preferably an Internet or UUCP host name) of the specific machine used
Like, imagine sending a benchmark score with zero context. 💀 Nobody knows what machine it came from, nobody knows what release ran it, and now the number means nothing. Don't do that to them. Give them the whole picture — every field above, fr fr.

And that's it! That's the whole assignment. Benchmark your box, scoop up the "summary" file, throw in all those delicious hardware and software deets, and send it off. Slay 💅

okay bestie, that's the entire page! we made it all the way to the end together 🎀

This was the BRL-CAD benchmark page — BENCHMARK(5) from the BRL-CAD manual, and the footer down at the bottom got a lil' truncated mid-title (BRL-CAD BENCHM...ITE METHODOLOGY(5), rip 💀), along with its date, so peep your own copy for the exact version-and-date stamp. Go forth, run the benchmark, and may your numbers be highkey iconic, your methodology correct, and your "summary" files always e-mailed to the developers 🔥👑

bye bestie, ily, hope your walrus results ate 🌈✨

📦 where this page comes from