Skip to content

QOCO-GPU Draft #29

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Closed
govindchari wants to merge 6 commits into from
Closed

QOCO-GPU Draft #29

govindchari wants to merge 6 commits into from

Conversation

@govindchari
Copy link
Member

@govindchari govindchari commented Nov 19, 2025

No description provided.

@govindchari
Replaced with QOCOMatrix and Vector
1bff246
@govindchari
Second attempt
0b64fc2
@govindchari
Checkpoint: Ruiz equilibration working with abstract types
f4cf4b2 
- Added abstract wrapper functions for row_inf_norm, col_inf_norm_USymm, and row_col_scale
- Uncommented and updated ruiz_equilibration to use QOCOMatrix* and QOCOVectorf*
- Fixed equilibration to match original implementation logic
- Re-enabled ruiz_test and lcvx_bad_scaling_test (both passing)
- update_vector_data_test still needs fixing
@govindchari
ix
a6cb19e
@govindchari
Use col_norm in ruiz
a0092dd
@github-actions
Copy link

github-actions bot commented Nov 19, 2025

Download benchmark artifacts

Benchmark Summary

  • Baseline solved: 129 problems
  • Diff branch solved: 8 problems

Differences in solved problems

  • Baseline solved additional problems: AUG2DCQP, AUG2DQP, AUG3DCQP, AUG3DQP, BOYD1, BOYD2, CONT-050, CONT-100, CONT-101, CONT-200, CONT-201, CONT-300, CVXQP1_M, CVXQP1_S, CVXQP2_L, CVXQP2_M, CVXQP2_S, CVXQP3_M, CVXQP3_S, DTOC3, DUAL1, DUAL2, DUAL3, DUAL4, DUALC1, DUALC2, DUALC5, DUALC8, EXDATA, GOULDQP2, GOULDQP3, HS118, HS21, HS268, HS35, HS35MOD, HS53, HS76, HUES-MOD, HUESTIS, KSIP, LASER, LISWET2, LISWET3, LISWET4, LISWET5, LISWET6, LISWET7, LOTSCHD, MOSARQP1, MOSARQP2, POWELL20, PRIMAL1, PRIMAL2, PRIMAL3, PRIMAL4, PRIMALC1, PRIMALC2, PRIMALC5, PRIMALC8, Q25FV47, QADLITTL, QAFIRO, QBANDM, QBEACONF, QBORE3D, QBRANDY, QCAPRI, QE226, QETAMACR, QFFFFF80, QFORPLAN, QGFRDXPN, QGROW15, QGROW22, QGROW7, QISRAEL, QPCBLEND, QPCBOEI1, QPCBOEI2, QPCSTAIR, QPILOTNO, QPTEST, QRECIPE, QSC205, QSCAGR25, QSCAGR7, QSCFXM1, QSCFXM2, QSCFXM3, QSCORPIO, QSCRS8, QSCSD1, QSCSD6, QSCSD8, QSCTAP1, QSCTAP2, QSCTAP3, QSEBA, QSHARE1B, QSHARE2B, QSHELL, QSHIP04L, QSHIP04S, QSHIP08L, QSHIP08S, QSHIP12L, QSHIP12S, QSIERRA, QSTAIR, QSTANDAT, S268, STADAT1, STADAT2, STADAT3, STCQP1, STCQP2, TAME, UBH1, VALUES, ZECEVIC2

@github-actions
Copy link

github-actions bot commented Nov 19, 2025

Download benchmark artifacts

Benchmark Summary

  • Baseline solved: 129 problems
  • Diff branch solved: 129 problems

Runtime regressions (> 5.0%)

  • AUG2D: diff=0.1295s, baseline=0.0105s, Δ=+1137.0%
  • AUG2DC: diff=0.1518s, baseline=0.0091s, Δ=+1564.6%
  • AUG2DCQP: diff=0.3065s, baseline=0.1209s, Δ=+153.6%
  • AUG2DQP: diff=0.3077s, baseline=0.1234s, Δ=+149.3%
  • AUG3D: diff=0.0165s, baseline=0.0016s, Δ=+962.1%
  • AUG3DC: diff=0.0161s, baseline=0.0016s, Δ=+934.6%
  • AUG3DCQP: diff=0.0545s, baseline=0.0208s, Δ=+162.3%
  • AUG3DQP: diff=0.0644s, baseline=0.0254s, Δ=+153.8%
  • BOYD1: diff=2.3705s, baseline=1.4087s, Δ=+68.3%
  • BOYD2: diff=4.2023s, baseline=2.3996s, Δ=+75.1%
  • CONT-050: diff=0.1043s, baseline=0.0529s, Δ=+97.0%
  • CONT-100: diff=0.6585s, baseline=0.4571s, Δ=+44.1%
  • CONT-101: diff=0.6793s, baseline=0.4903s, Δ=+38.6%
  • CONT-200: diff=5.6301s, baseline=4.8994s, Δ=+14.9%
  • CONT-201: diff=6.5110s, baseline=5.8087s, Δ=+12.1%
  • CONT-300: diff=22.9046s, baseline=21.1186s, Δ=+8.5%
  • CVXQP1_M: diff=0.0939s, baseline=0.0677s, Δ=+38.8%
  • CVXQP1_S: diff=0.0025s, baseline=0.0005s, Δ=+415.4%
  • CVXQP2_M: diff=0.0650s, baseline=0.0468s, Δ=+38.8%
  • CVXQP2_S: diff=0.0024s, baseline=0.0004s, Δ=+501.7%
  • CVXQP3_M: diff=0.1566s, baseline=0.1310s, Δ=+19.6%
  • CVXQP3_S: diff=0.0024s, baseline=0.0006s, Δ=+279.1%
  • DPKLO1: diff=0.0009s, baseline=0.0001s, Δ=+587.7%
  • DTOC3: diff=0.0945s, baseline=0.0116s, Δ=+714.6%
  • DUAL1: diff=0.0035s, baseline=0.0020s, Δ=+70.5%
  • DUAL2: diff=0.0038s, baseline=0.0023s, Δ=+67.5%
  • DUAL3: diff=0.0053s, baseline=0.0035s, Δ=+54.0%
  • DUAL4: diff=0.0024s, baseline=0.0013s, Δ=+85.1%
  • DUALC1: diff=0.0025s, baseline=0.0012s, Δ=+109.0%
  • DUALC2: diff=0.0022s, baseline=0.0008s, Δ=+168.0%
  • DUALC5: diff=0.0023s, baseline=0.0007s, Δ=+247.3%
  • DUALC8: diff=0.0042s, baseline=0.0015s, Δ=+176.3%
  • GENHS28: diff=0.0001s, baseline=0.0000s, Δ=+1675.0%
  • GOULDQP2: diff=0.0116s, baseline=0.0021s, Δ=+441.5%
  • GOULDQP3: diff=0.0139s, baseline=0.0015s, Δ=+833.6%
  • HS118: diff=0.0007s, baseline=0.0001s, Δ=+776.2%
  • HS21: diff=0.0001s, baseline=0.0000s, Δ=+691.7%
  • HS268: diff=0.0001s, baseline=0.0000s, Δ=+626.7%
  • HS35: diff=0.0001s, baseline=0.0000s, Δ=+766.7%
  • HS35MOD: diff=0.0001s, baseline=0.0000s, Δ=+394.4%
  • HS51: diff=0.0001s, baseline=0.0000s, Δ=+2066.7%
  • HS52: diff=0.0001s, baseline=0.0000s, Δ=+2900.0%
  • HS53: diff=0.0002s, baseline=0.0000s, Δ=+1120.0%
  • HS76: diff=0.0001s, baseline=0.0000s, Δ=+720.0%
  • HUES-MOD: diff=0.1085s, baseline=0.0274s, Δ=+295.7%
  • HUESTIS: diff=0.0955s, baseline=0.0158s, Δ=+505.0%
  • KSIP: diff=0.0120s, baseline=0.0074s, Δ=+63.1%
  • LASER: diff=0.0162s, baseline=0.0033s, Δ=+393.5%
  • LISWET2: diff=0.1149s, baseline=0.0209s, Δ=+449.8%
  • LISWET3: diff=0.1074s, baseline=0.0208s, Δ=+417.2%
  • LISWET4: diff=0.1181s, baseline=0.0230s, Δ=+414.0%
  • LISWET5: diff=0.1145s, baseline=0.0189s, Δ=+506.4%
  • LISWET6: diff=0.1156s, baseline=0.0209s, Δ=+453.7%
  • LISWET7: diff=0.1212s, baseline=0.0250s, Δ=+384.5%
  • LOTSCHD: diff=0.0003s, baseline=0.0000s, Δ=+634.3%
  • MOSARQP1: diff=0.0306s, baseline=0.0084s, Δ=+262.8%
  • MOSARQP2: diff=0.0160s, baseline=0.0061s, Δ=+163.0%
  • POWELL20: diff=0.1823s, baseline=0.0855s, Δ=+113.2%
  • PRIMAL1: diff=0.0060s, baseline=0.0043s, Δ=+40.4%
  • PRIMAL2: diff=0.0084s, baseline=0.0055s, Δ=+51.8%
  • PRIMAL3: diff=0.0175s, baseline=0.0139s, Δ=+26.3%
  • PRIMAL4: diff=0.0140s, baseline=0.0071s, Δ=+95.9%
  • PRIMALC1: diff=0.0054s, baseline=0.0023s, Δ=+135.1%
  • PRIMALC2: diff=0.0040s, baseline=0.0019s, Δ=+116.9%
  • PRIMALC5: diff=0.0051s, baseline=0.0021s, Δ=+144.9%
  • PRIMALC8: diff=0.0090s, baseline=0.0039s, Δ=+129.4%
  • Q25FV47: diff=0.2915s, baseline=0.2630s, Δ=+10.8%
  • QADLITTL: diff=0.0018s, baseline=0.0003s, Δ=+410.9%
  • QAFIRO: diff=0.0009s, baseline=0.0001s, Δ=+517.0%
  • QBANDM: diff=0.0098s, baseline=0.0046s, Δ=+112.0%
  • QBEACONF: diff=0.0057s, baseline=0.0025s, Δ=+128.5%
  • QBORE3D: diff=0.0060s, baseline=0.0022s, Δ=+174.4%
  • QBRANDY: diff=0.0066s, baseline=0.0033s, Δ=+98.9%
  • QCAPRI: diff=0.0135s, baseline=0.0088s, Δ=+54.6%
  • QE226: diff=0.0086s, baseline=0.0050s, Δ=+72.4%
  • QETAMACR: diff=0.0984s, baseline=0.0874s, Δ=+12.5%
  • QFFFFF80: diff=0.1198s, baseline=0.1098s, Δ=+9.1%
  • QFORPLAN: diff=0.0123s, baseline=0.0074s, Δ=+65.3%
  • QGFRDXPN: diff=0.0197s, baseline=0.0062s, Δ=+221.1%
  • QGROW15: diff=0.0182s, baseline=0.0085s, Δ=+114.6%
  • QGROW22: diff=0.0308s, baseline=0.0152s, Δ=+102.4%
  • QGROW7: diff=0.0091s, baseline=0.0043s, Δ=+114.1%
  • QISRAEL: diff=0.0063s, baseline=0.0038s, Δ=+67.3%
  • QPCBLEND: diff=0.0018s, baseline=0.0006s, Δ=+214.2%
  • QPCBOEI1: diff=0.0169s, baseline=0.0081s, Δ=+108.6%
  • QPCBOEI2: diff=0.0076s, baseline=0.0036s, Δ=+110.6%
  • QPCSTAIR: diff=0.0179s, baseline=0.0097s, Δ=+84.3%
  • QPILOTNO: diff=0.1650s, baseline=0.1281s, Δ=+28.8%
  • QPTEST: diff=0.0001s, baseline=0.0000s, Δ=+844.4%
  • QRECIPE: diff=0.0050s, baseline=0.0011s, Δ=+341.0%
  • QSC205: diff=0.0056s, baseline=0.0009s, Δ=+503.0%
  • QSCAGR25: diff=0.0120s, baseline=0.0022s, Δ=+434.4%
  • QSCAGR7: diff=0.0026s, baseline=0.0006s, Δ=+360.6%
  • QSCFXM1: diff=0.0126s, baseline=0.0063s, Δ=+98.7%
  • QSCFXM2: diff=0.0279s, baseline=0.0168s, Δ=+66.4%
  • QSCFXM3: diff=0.0392s, baseline=0.0254s, Δ=+54.2%
  • QSCORPIO: diff=0.0073s, baseline=0.0017s, Δ=+343.1%
  • QSCRS8: diff=0.0235s, baseline=0.0075s, Δ=+211.9%
  • QSCSD1: diff=0.0098s, baseline=0.0025s, Δ=+300.3%
  • QSCSD6: diff=0.0188s, baseline=0.0057s, Δ=+231.0%
  • QSCSD8: diff=0.0373s, baseline=0.0095s, Δ=+292.8%
  • QSCTAP1: diff=0.0103s, baseline=0.0024s, Δ=+334.2%
  • QSCTAP2: diff=0.0415s, baseline=0.0123s, Δ=+236.7%
  • QSCTAP3: diff=0.0570s, baseline=0.0184s, Δ=+210.0%
  • QSEBA: diff=0.0284s, baseline=0.0091s, Δ=+212.5%
  • QSHARE1B: diff=0.0059s, baseline=0.0021s, Δ=+184.1%
  • QSHARE2B: diff=0.0024s, baseline=0.0008s, Δ=+191.2%
  • QSHELL: diff=0.1652s, baseline=0.1358s, Δ=+21.6%
  • QSHIP04L: diff=0.0372s, baseline=0.0068s, Δ=+448.2%
  • QSHIP04S: diff=0.0262s, baseline=0.0047s, Δ=+451.0%
  • QSHIP08L: diff=0.2144s, baseline=0.1620s, Δ=+32.4%
  • QSHIP08S: diff=0.0631s, baseline=0.0278s, Δ=+127.2%
  • QSHIP12L: diff=0.3975s, baseline=0.3230s, Δ=+23.1%
  • QSHIP12S: diff=0.0777s, baseline=0.0380s, Δ=+104.3%
  • QSIERRA: diff=0.0709s, baseline=0.0188s, Δ=+276.3%
  • QSTAIR: diff=0.0195s, baseline=0.0120s, Δ=+62.1%
  • QSTANDAT: diff=0.0224s, baseline=0.0072s, Δ=+212.7%
  • S268: diff=0.0001s, baseline=0.0000s, Δ=+713.3%
  • STADAT1: diff=0.0699s, baseline=0.0157s, Δ=+344.3%
  • STADAT2: diff=0.0625s, baseline=0.0135s, Δ=+363.6%
  • STADAT3: diff=0.1234s, baseline=0.0257s, Δ=+381.0%
  • STCQP1: diff=0.1119s, baseline=0.0555s, Δ=+101.6%
  • STCQP2: diff=0.1986s, baseline=0.1305s, Δ=+52.3%
  • TAME: diff=0.0001s, baseline=0.0000s, Δ=+771.4%
  • UBH1: diff=0.2187s, baseline=0.0289s, Δ=+655.5%
  • VALUES: diff=0.0050s, baseline=0.0020s, Δ=+156.1%
  • ZECEVIC2: diff=0.0001s, baseline=0.0000s, Δ=+763.6%

@govindchari govindchari changed the title QOCO-GPU QOCO-GPU Draft Nov 19, 2025
@github-actions
Copy link

github-actions bot commented Nov 19, 2025

Download benchmark artifacts

Benchmark Summary

  • Baseline solved: 129 problems
  • Diff branch solved: 129 problems

Runtime regressions (> 5.0%)

  • BOYD2: diff=2.4506s, baseline=2.3217s, Δ=+5.6%
  • CVXQP2_L: diff=24.5875s, baseline=22.8309s, Δ=+7.7%
  • CVXQP2_M: diff=0.0497s, baseline=0.0465s, Δ=+6.9%
  • CVXQP2_S: diff=0.0004s, baseline=0.0004s, Δ=+9.3%
  • HS21: diff=0.0000s, baseline=0.0000s, Δ=+158.3%
  • HS268: diff=0.0000s, baseline=0.0000s, Δ=+13.3%
  • HS35: diff=0.0000s, baseline=0.0000s, Δ=+11.1%
  • HS53: diff=0.0000s, baseline=0.0000s, Δ=+6.3%
  • HS76: diff=0.0000s, baseline=0.0000s, Δ=+6.7%
  • PRIMALC8: diff=0.0040s, baseline=0.0038s, Δ=+5.6%
  • QAFIRO: diff=0.0001s, baseline=0.0001s, Δ=+5.3%
  • QPTEST: diff=0.0000s, baseline=0.0000s, Δ=+190.0%
  • QSCAGR7: diff=0.0006s, baseline=0.0006s, Δ=+6.7%
  • S268: diff=0.0000s, baseline=0.0000s, Δ=+6.3%
  • TAME: diff=0.0000s, baseline=0.0000s, Δ=+14.3%

Runtime improvements (> 5.0%)

  • CVXQP3_S: diff=0.0007s, baseline=0.0007s, Δ=-8.4%
  • DPKLO1: diff=0.0001s, baseline=0.0002s, Δ=-15.4%
  • GENHS28: diff=0.0000s, baseline=0.0000s, Δ=-16.7%
  • HS118: diff=0.0001s, baseline=0.0001s, Δ=-20.0%
  • HS35MOD: diff=0.0000s, baseline=0.0000s, Δ=-50.0%

@govindchari govindchari marked this pull request as draft November 19, 2025 18:38
@govindchari govindchari closed this Dec 4, 2025
@govindchari govindchari deleted the gc/qoco-gpu branch December 4, 2025 22:09
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@govindchari