CodeUniverse · package
QuSpin
many-Hamiltonian-entry gradient of a scalar observable
AD speed-up on the same differentiation workload (against finite differences) Speed-up: 125.9×
High-dimensional demo
many-Hamiltonian-entry gradient of a scalar observable
Supplied demo scenario wording.
FD components requested 1 – 256 · AD components 1024
View demo measurements
| Measure | Value |
|---|---|
| AD median (s) | 0.0005918750539422035 |
| FD median (s) | 0.07450524996966124 |
| FD / AD | 125.8800307149567 |
| Max scaled error | 3.594327504900796e-11 |
| AD components | 1024 |
| FD components requested | 256 |
| AD parameters | 256 |
| Repeats | 5 |
| FD components requested | AD median (s) | FD median (s) | FD / AD | Max scaled error | AD components |
|---|---|---|---|---|---|
| 1 | 1024 | 0.0005930420011281967 | 0.00030045793391764164 | 0.5066385405183 | 3.886691966534239e-12 |
| 2 | 1024 | 0.0005882089026272297 | 0.0005997910629957914 | 1.0196905560538614 | 1.847927648575276e-11 |
| 4 | 1024 | 0.0006013750098645687 | 0.0011767499381676316 | 1.9567656102514783 | 3.07724434042006e-11 |
| 8 | 1024 | 0.0005898750387132168 | 0.0023678329307585955 | 4.014126340934693 | 3.07724434042006e-11 |
| 16 | 1024 | 0.0005792919546365738 | 0.004697541007772088 | 8.109107972540635 | 3.07724434042006e-11 |
| 32 | 1024 | 0.0005684170173481107 | 0.009213500074110925 | 16.209050385394043 | 3.284290135859014e-11 |
| 64 | 1024 | 0.000604208093136549 | 0.018473000032827258 | 30.573903664432418 | 3.284290135859014e-11 |
| 128 | 1024 | 0.000593916978687048 | 0.0373947499319911 | 62.962924573495776 | 3.284290135859014e-11 |
| 256 | 1024 | 0.0005918750539422035 | 0.07450524996966124 | 125.8800307149567 | 3.594327504900796e-11 |
One-direction performance audit
View audit measurements
| Demo scenario | Problem size | Repeats | AD median (s) | FD median (s) | FD / AD | Relative error | Supplied verdict |
|---|---|---|---|---|---|---|---|
| many-Hamiltonian-entry gradient of a scalar observable | — | — | 0.03248375 | 0.00206337508 | 0.0635202 | — | — |
| many-Hamiltonian-entry gradient of a scalar observable | — | — | 0.00627391599 | 0.00153916702 | 0.245328 | — | — |
| many-Hamiltonian-entry gradient of a scalar observable | — | — | 0.00183058297 | 0.00126658299 | 0.691900947 | — | — |
| many-Hamiltonian-entry gradient of a scalar observable | — | — | 0.003405082971 | 0.002745624981 | 0.806331 | — | — |
| many-Hamiltonian-entry gradient of a scalar observable | — | — | 0.002297417028 | 0.001963333925 | 0.854583 | — | — |
| many-Hamiltonian-entry gradient of a scalar observable | — | — | 0.001516000018 | 0.001315417001 | 0.867689 | — | — |
| many-Hamiltonian-entry gradient of a scalar observable | — | — | 0.001726375078 | 0.001512625022 | 0.876186 | — | — |
FD/AD > 1 means automatic differentiation is faster; FD/AD < 1 means the finite-difference comparison is faster. One direction only, measured inside the package's own case.