General description

As of May 2024, CAMELS contains 14,037 cosmological simulations: 6,136 N-body and 7,901 hydrodynamic simulations.

All simulations follow the evolution of \(256^3\) dark matter particles and \(256^3\) initial gas resolution elements (only for the hydrodynamic) within a periodic comoving volume of \((25~h^{-1}{\rm Mpc})^3\) from \(z=127\) down to \(z=0\).

Note

CAMELS is expanding and will soon include simulations with \(512^3\) dark matter particles and \(512^3\) gas resolution elements in periodic boxes of \((50~h^{-1}{\rm Mpc})^3\). CAMELS has recently expanded to include a set of zoom-in simulations of massive halos, which are centered in periodic boxes of \((200~h^{-1}{\rm Mpc})^3\).

For each simulation, we store multiple snapshots. We also keep a variety of post-processed data such as halo catalogs, power spectra…etc; see Data organization.

Organization

The simulations can be classified into suites, volumes, and sets, depending on the code used to run them, their volume, and how their parameters are organized. See Organization for further details. The table below shows the number of available hydrodynamic and N-body simulations:

Hydrodynamic simulations
	Set
Suite	SB	LH	1P	CV	EX	BE	Total
IllustrisTNG	2048 (28 params)	1000 (6 params)	113 (28 params)	27	4	27	3219
SIMBA		1000 (6 params)	113 (28 params)	27	4	27	1171
Astrid	1024 (7 params)	1000 (6 params)	25 (6 params)	27	4		2080
Magneticum		50 (6 params)		27			77
Swift-EAGLE		1000 (6 params)	25 (6 params)	27			1052
Ramses	256 (5 params)	–	13 (5 params)	27			296
Enzo		–	5	1			6
All	3328	4050	294	163	12	54	7901

N-body simulations
	Set
Suite	SB	LH	1P	CV	EX	BE	Total
IllustrisTNG	2048 (5 params)	1000 (2 params)	9 (5 params)	27	1	27	3112
SIMBA		1000 (2 params)					1064
Astrid	1024 (3 params)	1000 (2 params)					2088
Magneticum							64
Swift-EAGLE							64
Ramses		–					64
Enzo		–					64
All	3072	3000	9	27	1	27	6136

Note

CAMELS is designed so that each hydrodynamic simulation has a N-body counterpart. However, several hydrodynamic simulations may have the same N-body simulation. For instance, in the 1P set, simulations that only vary astrophysical parameters, will have the same N-body counterpart. This is the reason why in the above table, the number of N-body simulations may be lower than the corresponding number of hydrodynamic simulations.

The value of the cosmological, astrophysical, and initial random seed of the different simulations can be found Parameters.

Codes

All N-body simulations have been run with the Gadget-III code, while the hydrodynamic simulations have been run with different codes: AREPO, GIZMO, MP-Gadget, OpenGadget, Swift, Ramses, and Enzo. See Codes for details on the different codes and subgrid physics models available in CAMELS.

Redshifts

Important

A reorganization of the data has been performed in 2024 in order to enhance its uniformity and simplicity. This will require slight changes to existing codes that access the data.

• Snapshot numbers in the IllustrisTNG and SIMBA suites, where simulations have only 34 snapshots, have been updated to match the numbering in the Astrid suite (and some TNG simulations) that have 91 snapshots. For example, where 33 used to be the z=0 snapshot, now it is 90 uniformly for all suites.

All simulations have 91 snapshots. This file contains the scale factors of the snapshots. The user can click below to see the redshift, scale factor, and snapshot number for the different snapshots:

Click here to see the relation between the snapshot number and the redshift

Redshift	Scale Factor	34 snapshots	91 snapshots
14.99	0.063	—	000
13.34	0.070	—	001
11.98	0.077	—	002
11.20	0.082	—	003
10.48	0.087	—	004
9.69	0.094	—	005
9.00	0.100	—	006
8.49	0.105	—	007
8.01	0.111	—	008
7.60	0.116	—	009
7.24	0.121	—	010
6.89	0.127	—	011
6.56	0.132	—	012
6.28	0.137	—	013
6.01	0.143	000	014
5.75	0.148	—	015
5.50	0.154	—	016
5.23	0.161	—	017
5.00	0.167	001	018
4.80	0.172	—	019
4.61	0.178	—	020
4.45	0.183	—	021
4.30	0.189	—	022
4.15	0.194	—	023
4.01	0.200	002	024
3.87	0.205	—	025
3.73	0.211	—	026
3.62	0.216	—	027
3.49	0.223	003	028
3.36	0.229	—	029
3.24	0.236	—	030
3.12	0.242	—	031
3.01	0.249	004	032
2.90	0.257	—	033
2.80	0.263	005	034
2.72	0.269	—	035
2.63	0.276	006	036
2.54	0.282	—	037
2.46	0.289	007	038
2.38	0.296	—	039
2.30	0.303	008	040
2.22	0.310	—	041
2.15	0.318	009	042
2.07	0.325	—	043
2.00	0.333	010	044
1.93	0.341	—	045
1.86	0.349	011	046
1.80	0.358	—	047
1.73	0.366	012	048
1.67	0.375	—	049
1.60	0.384	013	050
1.54	0.393	—	051
1.48	0.403	014	052
1.43	0.412	—	053
1.37	0.422	015	054
1.30	0.434	—	055
1.26	0.443	016	056
1.20	0.455	—	057
1.14	0.466	017	058
1.09	0.478	—	059
1.05	0.489	018	060
1.00	0.501	—	061
0.95	0.513	019	062
0.90	0.525	—	063
0.86	0.538	020	064
0.82	0.550	—	065
0.77	0.564	021	066
0.73	0.577	—	067
0.69	0.591	022	068
0.65	0.605	—	069
0.61	0.620	023	070
0.58	0.635	—	071
0.54	0.650	024	072
0.50	0.665	—	073
0.47	0.681	025	074
0.43	0.698	—	075
0.40	0.714	026	076
0.37	0.731	—	077
0.34	0.749	027	078
0.30	0.771	—	079
0.27	0.789	028	080
0.24	0.808	—	081
0.21	0.827	029	082
0.18	0.847	—	083
0.15	0.867	030	084
0.13	0.888	—	085
0.10	0.910	031	086
0.07	0.931	—	087
0.05	0.954	032	088
0.02	0.977	—	089
0.00	1.000	033	090

Note

The exact redshifts of a given snapshot may be slighty different to the above ones. For instance, there are small differences between the exact redshifts of the snapshots in the Astrid and SIMBA suites. In the simulations of the IllustrisTNG suite these numbers can also be slightly different, since AREPO can only write snapshots in the highest time steps in the hierarchy.

Warning

We have moved some snapshots to tape. If you need these please reach out.