Int32 Overflow in whole Hilbert-space

[del2341]

Thank you for last reply. I have another question. It appears that the current version of HPhi only supports basis sizes that fit within a 32-bit integer. Is there any way to continue calculations when the whole Hilbert-space dimension exceeds the Int32 limit? For example, consider the following setup. The final dimension is just 45:

W = 15
L = 3
model = "FermionHubbard"
method = "CG"
lattice = "triangular"
t = 1.0
U = 4.0
nelec = 46
2Sz = 44

I will meet with an error as "Error: Fail for memory allocation. see error code 1 in user manual.".
Any guidance would be greatly appreciated.

[tmisawa]

Unfortunately, the HPhi code is written assuming 64-bit integers, so this case cannot be handled.
In particular, some bitwise operations (such as snoob) are based on the assumption that the integer is 64-bit.

In principle, it would be possible to handle Hubbard models with more than 32 sites
by splitting the integer into segments smaller than 64 bits.
However, implementing that in HPhi would require a significant rewrite, so we can not do that immediately.

[tmisawa]

For your information, in the case of spin-1/2 systems, there is a software package called QS3 that performs Lanczos calculations near the saturated magnetization sector of Sz (see the following link).
By using algorithms implemented in QS3, it might be possible to perform calculations for the Hubbard model.

https://github.com/QS-Cube/ED

[del2341]

Thank you so much for your quick response. I will try that.