Mikhail,
Yes, we've made significant progress on reviving the GPU version. If you download the latest release, you'll find it included. Feedback is welcome.
--Michael
Search found 2065 matches
- 02 Nov 2026, 17:38
- Forum: Feature Requests
- Topic: Development of GPU version of elegant
- Replies: 4
- Views: 3493
- 02 Nov 2026, 11:33
- Forum: Ring Tracking
- Topic: Transverse bunch-by-bunch feedback simulation
- Replies: 1
- Views: 1845
Re: Transverse bunch-by-bunch feedback simulation
Chao,
There's no way presently to do this in the transverse plane. It's a good idea though and I'll see what I can do to implement it. Doesn't sound difficult.
--Michael
There's no way presently to do this in the transverse plane. It's a good idea though and I'll see what I can do to implement it. Doesn't sound difficult.
--Michael
- 02 Nov 2026, 11:31
- Forum: Ring Tracking
- Topic: space charge solver
- Replies: 3
- Views: 2344
Re: space charge solver
Chao, That wouldn't be hard to add to the existing space charge algorithm. However, if the x-y tilt varies with s, it would introduce errors because elegant uses the ordinary (uncoupled) twiss parameters to provide a weighting factor for each space charge kick. It would presumably be accurate in the...
- 13 Oct 2026, 10:31
- Forum: Ring Tracking
- Topic: Top up tracking safety
- Replies: 1
- Views: 2228
Re: Top up tracking safety
Samira, Top-up (or swap-out) safety tracking is a very large simulation campaign and it is going to be quite difficult to provide a useful example. Our approach for the APS upgrade is summarized here . Perhaps that is a helpful starting point. Some low-energy rings (e.g., ALS-U) avoid this issue by ...
- 13 Oct 2026, 10:24
- Forum: Ring Tracking
- Topic: frequency map
- Replies: 4
- Views: 4866
Re: frequency map
Samira,
I'm not sure what the issue might be. We compared elegant and AT in the past for nonlinear dynamics and found good agreement. See
this paper.
--Michael
I'm not sure what the issue might be. We compared elegant and AT in the past for nonlinear dynamics and found good agreement. See
this paper.
--Michael
- 29 Sep 2026, 13:42
- Forum: Ring Tracking
- Topic: space charge solver
- Replies: 3
- Views: 2344
Re: space charge solver
Chao,
Because the space charge is applied as a kick at discrete locations, which involves integrals over the beta functions up and downstream of the kick location, it isn't straightforward to properly account for x-y tilts. At present, elegant doesn't attempt to do it.
--Michael
Because the space charge is applied as a kick at discrete locations, which involves integrals over the beta functions up and downstream of the kick location, it isn't straightforward to properly account for x-y tilts. At present, elegant doesn't attempt to do it.
--Michael
- 23 Sep 2026, 12:38
- Forum: Ring Tracking
- Topic: simulation with a (time-depended) bumper
- Replies: 6
- Views: 10846
Re: simulation with a (time-depended) bumper
Chao, Sorry for not answering sooner. Sometimes the forum doesn't email me when new questions appear. I think the simulation is working as I expect it to. If I compute the change in kick angle from the coordinate WATCH files, I get a curve with the expected shape and height sddsxref FANKICKWATCH_2.W...
- 23 Sep 2026, 10:14
- Forum: Ring Tracking
- Topic: frequency map
- Replies: 4
- Views: 4866
Re: frequency map
Samira, You can offset the x and y axes values by the integer tunes using the -offset option in sddsplot, e.g., sddsplot -col=nux,nuy fma_1.fma -graph=sym,vary=subtype,fill -split=col=diffusionRate -order=spect -offset=xch=7,ych=6 -scale=7.2,7.28,6,6.4 Looking at your lattice file, I noticed that yo...
- 01 Sep 2026, 12:23
- Forum: General
- Topic: Version 2026.4 released
- Replies: 0
- Views: 12369
Version 2026.4 released
Version 2026.4 has been released. The main improvement is addition of features that benefit long time-scale simulations, to avoid phase drifts due to rounding error. The manual has more details.
--Michael
--Michael
- 05 Aug 2026, 10:08
- Forum: Ring Tracking
- Topic: simulation with a (time-depended) bumper
- Replies: 6
- Views: 10846
Re: simulation with a (time-depended) bumper
The problem is related to the way the BUMPER element aligns the time coordinates of the beam to the time coordinates of the waveform. In particular, it computes <t> for the particles and aligns that to t=0 in the waveform file. If you center your waveform t data about t=0, you should get the results...